MSU LIBRARIES ”- RETURNING MATERIALS: ace n 00 rop to remove this checkout from your record. FINES wi11 be charged if book is returned after the date stamped be10w. A‘- CASE RELATIONSHIPS AND THE VERB MATRIX IN PERSIAN AND ENGLISH (A CONTRASTIVE STUDY) By Seyed-Ali Miremadi A DISSERTATION Submitted to Michigan State University if] pyartial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Linguistics and Oriental and African Languages 1981 tn. .4. ,1 u a ABSTRACT CASE RELATIONSHIPS AND THE VERB MATRIX IN PERSIAN AND ENGLISH By Seyed-Ali Miremadi Robert Longacre in 1976 attempted to catalogue the notional categories that, to him are "an impor- tant part of our cognitive/notional apparatus as human beings" . The scope of Longacre's study is universal. That is , he claims that the generalizations are valid relations between notional-superficial interactions for all languages. By using the case grammar model proposed by Longacre as the basis for classification, this dissertation concerns itself with the complex deep-surface relationships of Persian clauses, attempt- ing to 1) describe the case frame system of Persian, and 2) test Longacre's claim that his approach is universally applicable by providing a description of a language other than English. This study emphasizes case frames and predication relationships. An effort has been made to identify conflicts between the case frame systems of Persian and English, and to provide suggestions for future studies. The corpus upon which the study is based consists of 405 clauses taken from Daijan Napeleon by Iraj Pezeshkzad; other examples have been incorporated into the corpus where the data were lacking. .4 . . u. . . . .. . . u a. .- a... u. .U. \§. an. O V .. .,. .. ... .. .. .5. .. ~ .I ~. ~ I I. v a. Seyed-Ali Miremadi T113 study is divided into six chapters. Chapter one illttnaduces the problem and describes the methodo- logy <3f 'the study. A general review of the evolution of the: analysis of case in English is presented in this chaptenr. Chapter two presents a brief explanation of Longacre's case analysis system, as well as the nature of tagmemic theory, which constitutes the theoretical framework of this study. Chapter three covers cases and case frames in Persian. Predicates are shown to be composed of a number of complex features, which in turn determine the features of accompanying nominals. In Chapter four, a review of the literature concerning Iranian scholars' contributions to case analysis is presented, showing the disparities between their analyses and the present analysis. Chapter five briefly examines possible areas of conflicts between English and Persian. Chapter six concludes the study. The list of verbs analyzed constitutes the Appendix I of this study. The meaning and a case frame are assigned to each verb entry, followed by a typical example. Appendix II includes the sources to sentences used in Chapter three. To my wife Forough Miremadi ii ACKNOWLEDGEMENTS My sincere appreciation is extended to Dr. Ruth M. Brend, the chairperson of my doctoral committee. She devoted many hours to reading and editing this work, as well as providing encouragement and suggestions. I would also like to thank Dr. Julia Falk, professor of linguistics and Assistant Dean of College of Arts and Letters; Dr. Carol Scotton; and Dr. Paul Munsell, my committee members for their invaluable assistance and counsel during the preparation of this dissertation. I would like to note my appreciation to Dr. James Stalker, the College Representative who was kind enough to read the entire dissertation and offer valuable suggestions. A special note of thanks is reserved for Professor James Wang, the chairman of the Department of Linguistics and Oriental and African Languages; Dr. David G. Lockwood for providing suggestions on the section on Stratificational Grammar; and Dr. Barbara Abbott, for her patience in answer— ing my questions on linguistics in general during my four years of studies at Michigan State University. I would also like to thank Michael Eric Bennett, both a Ph.D. candidate in the Linguistic Department and a friend for his assistance in editing earlier drafts. I would like to express my gratitude to Mrs. Jo-Ann Vander Bergh iii for her encouragement and kindness; and Ms. Betty Briggs and Ms. Clara Hanna who spent many hours transforming my handwriting into the final product. Lastly, a special thanks to my wife, Forough and our children, Kamran, Mandana, and Kamyar; my brother, MORTEZA MIREMADI; and Judith Miremadi. It was their support, en- couragement, and sacrifice that made this project possible. iv TABLE OF CONTENTS Page List of Charts The Sounds of Persian Abbreviations Used in ths Work Chapter One 1.1. Objectives and Significance of this research... 1 1.2. The need for a classification.................. 3 1.3. The Problem.............. ................ . ..... 6 1.3.1. Hypotheses. ... . ..... . ......................... 8 1.4. Data Collection.. ....................... ....... 10 1.4.1. Procedures. ....... ........... .... ........... . 10 1.5.1. Historical Background ...... ............ . ...... 12 1.5.2. Contemporary Non- transformational Scholars... 15 1.5.2.1. de Groot....................................... 15 1.5.2.2. Halliday............................ ........... 18 1.5.3. The Contribution of Tagmemicists .......... ..... 22 1.5.3.1. Platt.. ....... . ........... . ................... . 22 1.5.3.2. Longacre.............................. ......... 25 1.5.3.3. Cook .............. . ............................ 25 1.5.4. Case Grammar and Transformational Grammar...... 28 1.5.4.1. Fillmore's Case Grammar ........................ 30 1.5.4.2. Chafe............................... ........... 34 1.5.4.3. Anderson ............................... . ....... 42 1.5.5. Case Grammar and Stratificational Theory....... 46 1.5.6. Case Analysis and Functional Grammar (Dik) ..... 49 Notes to Chapter One.. ..... .... ................ 53 Chapter Two Theoretical Framework 2.1. Longacre... ................................... . 56 2.1.1 Predication.... ................................ 59 2.1.2. Cases or Roles ................................. 60 2.1.2.1. Experiencer... ................................. 62 2.1.2.2. Patient ........................................ 63 2.1.2.3. Agent............ ......... . .................... 63 2.k.2.4. Range.... ...................................... 63 2.1.2.5. Measure..... .................. ..... ............ 64 2.1.2.6. Instrument ..................................... 64 2.1.2.7. Locative... ...... . .............. .. ............. 65 V 2.1.2.8. Source ......................................... 65 2.1.2.9. Goal ........................................... 66 2.1.2.10. Path ........................................... 67 2.2. Non-nuclear Cases .............................. 67 2.2.1. Peripheral Locative... .............. . .......... 67 2.2.2. Time, Manner, Cause, Purpose, Accompaniment.... 68 2.3. Parameters ..................................... 68 2.4. Deep and Surface Structure ..................... 72 2.5. Tagmemic as a Model... ......................... 75 Notes to Chapter Two ........................... 81 Chapter Three Case Analysis of Persian 3.1. Introduction ................................... 84 3.1.1. Definition of terms ............................ 87 3.1.2. Data....................... .................... 93 3.1.3. The Problems of Case Analysis........ .......... 94 3.1.4. Theoretical Assumptions ........................ 97 3.1.5. The Language Under Study ....................... 99 3.1.6. Word Order.... ................................. 99 3.1.7. Persian Verbs .................................. 101 3.1.7.1. Simple and Compound Verbs ...................... 101 3.1.7.2. Auxiliary Verbs....... ........... . ............. 105 3.2. Nominal Features. ....... . ........... . ........ .. 106 3.2.1. The Feature/Intentl. ......... . ................. 106 3.2.2. The Features/Instigator/and/Performer/ ......... 108 3.2.3. The Feature/Potent/ ............................ 111 3.3. Nuclear Cases.. ................................ 112 3.3.1. Direction Group... ....... . ..... . ............. .. 113 3.3.1.1. Experiencer... ......................... . ....... 114 3.3.1.2. Patient ....................... . ................ 118 3.3.1.3. Goal ........................................... 122 3.3.1.4. Client ......................................... 125 3.3.2. Agent .......................................... 127 3.3.3. Giver Group .................................... 131 3.3.3.1. Source.. ....................................... 131 3.3.3.2. Path......... .................................. 134 3.3.4. Instrument ..................................... 137 3.3.5. Locative ....................................... 140 3.3.6. Neutral Group .................................. 142 3.3.6.1. Range .......................................... 142 3.3.6.2. Measure..... ................................... 143 3.3.6.3 Identifier ..................................... 144 3.3.7. Peripheral Roles ............................... 145 3.4. Devices for Predication Categorization ......... 146 3.4.1. Physical ....................................... 146 3.4.2. Experiential ................................... 147 3.4.3. Ambient ..... . .................................. 147 3.4.4. Possession ..................................... 148 3.4.5. Directed ....................................... 148 vi afl\ I11. 3.4.6. Intentional ....... . ........................... 3.4.7. Completable......... ..... .. ................ ... 3.4.8. Measurable.......... ..... . ......... . ......... . 3.4.9. Locative.... .................................. 3.4.10. Motion...... .................................. 3.4.11. Instrumental.......... ........................ 3.5. Case Frames.. ....... ................... ...... . 3.5.1. Group A: Ambient Case Frames........ ....... .. 3.5.2. Group B: Ambient Experiential Case Frames.... 3.5.3. Group C: Experiential Case Frames ........... . 3.5.4. Group C': Factual Knowledge Case Frames ....... 3.5.5. Group D: Experiential Directed Case Frames... 3.5.6. Group D': Experiential Completable Directed Case Frames.............................. 3.5.7. Group E: Physical Case Frames................ 3.5.8. Group F: Measurable Case Frames..... ........ . 3.5.9. Group G: Locative Case Frames ............... . 3.5.10. Group G': Physical Motion Case Frames ......... 3.5.11. Group H: Physical Possession Directed Case Frames....................... ....... 3 5.12. Group H': Physical Possession Motion Case Frames........ .......... . ................ 3.5.13. Group I: Equative Case Frames ................ 3.6. Passive Constructions ..... ..... ..... . ......... 3.7. Reflexive Case Frames......................... 3.8. Surface Clause Structure Types (Tagmemic Formulas).......................... ...... 3.8.1. Clause C1ass..... ............................. 3.8.2. The Surface and Deep Structure of Persian Clauses. ..... . ..... . ........... . ........ . 3.9 Conclusion.................. ........ . ...... . Notes to Chapter Three.... ...... . ............. Chapter Four 4.0. Studies on Persian Grammar and Case Ana1y51s 4.1. Introduction....................... ...... ..... 4.2. Traditional Studies of Predication. ........... 4.2.1. William Jones..................... ........... . 4.2.2. Forbes....... ........ .. ...................... . 4.2.3. Phillott.......... ............................ 4.2.4. Levy. .......... . ................ . ............. 4.2.5. Homayun Farrokh...... ............... . ......... 4.3. Modern Non-transformational Grammarians of Persian........... ...... ...... ........... 4.3.1. Lazard.... ...................... . ............. 4.3.2. Bateni....... ................................. 4.3.3. Khanlari.... ...... . . ......................... 4.3.4. Lambton................. . .......... .. . ..... 4.4. Transformational Grammar and Persian ........ .. 4.4.1. Palmer........ ................. ............... 4.4.2. Marashi. ................................. . vii 150 150 151 151 151 152 152 155 158 161 164 167 170 175 177 180 183 185 186 190 197 201 203 204 217 218 4.4.3. Bashiri....................................... 247 4.5. Conclusion. ................................... 250 Notes to Chapter Four... ..... ...... ..... ...... 252 Chapter Five Comparison 5.0. Introduction.................................. 256 5.1. English Case Frames and Persian Case Frames: the Emergence of Differences............. 262 5.1.1. Ambient Case Frames.............. ........... .. 263 5.1.2. Ambient Experiential Case Frames.............. 264 5.1.3. Experiential (Completable) Case Frames. ...... . 265 5.1.4. Experiential Directed Completable Case Frames. 266 5.1.5. Physical Motion Case Frames................... 267 5.2. Word 0rder..................... . .............. 268 5.3. Superficial Morphological Differences......... 269 5.4. Cognate Prolongation.................... ...... 272 5.5. The Conflict in Lexicalized or Incorporated Verbs................ ............. ....... 274 5.6. Integrated Clauses..... ....... ................ 275 5.7. Conclusion....... ..... . ............ . ...... .... 276 Notes to Chapter Five..... ..... ............... 278 Chapter Six Summary, Limitation, Extension 6.1. Introduction................ .............. .... 279 6.2. Limitation............. ....................... 283 6.3. The Contribution of This Study................ 284 6.4 Suggestions for Further Studies ............... 286 6.5 Conclusion.................................... 287 Notes to Chapter Six.......................... 288 Appendix 1 Some Verb Entries in Persian ................. 289 Appendix 2 Data Base .................................... 321 Bibliography.. .......................................... 327 viii LIST OF CHARTS Page English Case Frames........ ........ . ........... ....... 69-71 Deep Level Versus Surface Levels... ................... 82 Verb Distribution Matrix.. ....... ..................... 94 Integrated Case Frames in Persian....... ......... ..... 189 The Surface and Deep Structure of Persian Clauses ..... 203-217 Underlying-surface Relationships in Persian and English......................... ..... ............ 258 ix u ‘ -\‘ «In THE SOUNDS OF PERSIAN A. CONSONANTS The consonant sounds of Persian with approximate indications of their values in terms of English sounds are as follows: p pir 'old' pack b him 'fear' beam t tir 'arrow' team k kif 'bag' 5111 g gur 'grave' go q 3am _ 'grief' f farman 'order' foot V 3a 'and' Village 3 Eib 'apple' gee z gud 'soon' £00 8 gab 'night' shoe 2 iarf 'deep' measure x gas 'happy' — c deém 'eye' ghurgh j iasn 'party' iudge 1 lab 'lip' lip r Luz 'day' river m mfih 'moon' moon n 35m 'name' name y yar 'friend' yard B. VOWELS The vowel sounds of Persian with approximate indi- cations of their values in terms of English sounds are as follows: i pir 'old' leave e sgr 'secret' bed a sar 'head' sad 5 bér 'load' car 0 bgz 'goat' u pul 'money' p331 Ach Amb int. loc/l inst CAUS I M Marg Intention NP poss phys Pron Prep ABBREVIATIONS USED IN THIS WORK Accusative Marker Ambient Agent Adjective Completable Ezafe Experiential Experiencer Goal intrasitive locative instrumental Causation Instrument Measure Margin Intentional Noun Phrase possession Physical Pronoun Preposition xi TV tr pred IND "d l \_.‘ ,_J\_.,_l\._v_l RH present progressive Patient Range Subjunctive Statment Source phrase Transitive Marker transitive predicate indicative unacceptable odd but acceptable State Case Frame Process Case Frame Action-Process Case Frame Action Case Frame xii Chapter One 1.0. Introduction 1.1. Objectives and Significance of this Research The purpose of this study is to categorize the predicates of Persian in such a manner as to specify the syntactic-semantic relationships which exist be— tween the predicate and other elements of the clause, which are centered around it. The classification of predications will be based on a semantico—syntactic relationship model recently preposed by Robert E. Longacre (1976). Longacre's model is a good corner- stone for case analysis, since he claims that his system is universally applicable. The present work will also test the value of Longacre's model in describing a language other than English. Two goals have been set for this study. The first is a linguistic analysis of case—frames and the case relationships in Persian; this will include the major verb categories and their related notions. The second is a contrastive analysis of the Persian case-frame system as opposed to the English case system. The outcome of this study will demonstrate the similarities as well as»‘the differences between the two systems. The analysis will establish a foundation for later studies of Persian grammar. It may also make some contribution to later studies on language universals. The general objective will be to explore eXperi- mentally and systematically certain current theoretical ideas dealing with the organization and arrangement of case-frames within the linguistic systems of Persian and English. I will particularly focus on how case frames in English in general and in Persian in parti- cular appear on the surface. I am primarily interested in the following: a. describing and developing the case-frame system of English as it has been interpreted and analyzed by Robert Longacre in his book An Anatomy of Speech Notions (1976).1 This analysis will be used as a qualitative reference against which a similar analysis of Persian will be made; b. describing the case—frames of Persian and English with the intention of making a contras- tive analysis. This analysis will highlight the differences which exist between the two languages; and ‘1‘. c. illustrating the manner in which the deep structures provided by the case analysis are transformed (or in Longacre's term 'elaborated') into surface structures in Persian. Understanding the message conveyed by a sentence requires knowledge of many factors, most of which will not be focused upon in this research. Factors such as cohesion within the context--though crucial in a more complete analysis--methaphoric expression and ellipsis, and so on will be excluded. However, the impact on meaning of permutations within the Persian surface clause structure will be taken into account. 1.2. The need for a classification. Despite the disparity in their surface realizations in different languages, features characterizing under- lying representations tend to be more similar than different. Although most languages have not been fully described, the available data tend to support the hypotheses of the universality of the existence of case relations in all languages. Fillmore says: "I believe that human languages are con- strained in such a way that the relations between arguments and predicates fall into a small number of types. ...these role types can be identified with certain quite elementary judgments about the things that go on around us: judgments about who does something, who experiences something, who benefits from something, where something happens, what it is that changes, what it is that moves, where it starts out, and where it ends up" (1968b: 382). Language is the most accessible and at the same time the most sophisticated means through which one can intrude into others' minds to extract their psychologi~ cal concepts, which might be universal. In other words, the observable linguistic elements are the manifestations and reflections of thought. The classification of case frames (i.e., verbs + cases (roles))is a significant step towards understanding "language universals". The attempt to discover verb-noun relationships is more important as an intermediate step than as an end in itself. As Longacre states, "The motivation for such a classification [i.e., case frame analysis] ... is not that taxonomy is an end in itself, but that such a taxo— nomy is useful" (1976:38). It is useful in the sense that it can make further generalizations possible. (For the remainder of this chapter all references to Longacre will be to his 1976 study, unless otherwise indicated). This may imply that taxonomy by itself is a milestone to more sophisticated analyses of language- thought relationships. In the past two decades, linguistics has become significantly more demanding with regard to the how's and why's of the grammatical elements within the propositions. If one needs to discover how speakers of a language manage to transmit information, and moreover, how they understand utterances, it is not enough to deal with surface structures in which linguis- tic elements occur. Deep or semantic structures should be fully developed and analyzed in order for one to appreciate the deep-surface relationships which facili- tate communication among the members of a linguistic community. Although no classification based on a single language is perfect enough to be thoroughly applicable to the analyses of other languages, any classification of single language data is by itself a step towards understanding language universals. A clear and understandable classification of the case-frame system of a language requires that the investigator: a. discover and specify the features which are attributed to verbs, b. discover and specify the cases (roles) which are characteristically related to verbs, c. make generalizations and categorizations of the features which distinguish one case-frame from another, and d. discover and clarify the nature of semantic relationships which exist between verbs and cases (roles) which occur with them. An understanding of the case-frames of a language is pertinent to the comparison of two or more languages. The comparison, as an end by itself, is useful. However, it is more useful if it contributes to an understanding of similarities and dissimilarities among universal semantic primitives. It also helps discover how the differences in semantic-Syntactic links differentiate one particular language from other languages. 1.3. The Problem Case grammar has been used with different linguistic models. Platt's tagmemic analysis (1971), and Jacken- doff's interpretive semantic analysis (1972) provide examples. If one narrows the scope of semantic analysis to the realm of single predication (Longacre), one will notice that a case grammar, if developed systematically, has the potential of relating one semantic structure to various surface structures and vice-versa. Moreover, there are cases where syntax by itself cannot explicitly demonstrate relationships required to explain paraphrase and ambiguity (Cook 1979:VI). Longacre states that any specification of surface structure serving to describe an unanlyzed language requires an inventory of roles or cases. Such an inventory will also help if one needs to understand a language in its own terms or to explore other languages (27). It has also been claimed that "...the deep notional categories ... are an important part of our cognitive apparatus as human beings" (Longacre:20). The idea that notional categories are in the underlying structures of all languages is not a new one. The universality of notional categories and the verbs with which they are associated has been a matter of deliberation (see Lyons 1966). Taking into consideration all the suggestions made concerning universal underlying cases in all languages, it is reasonable to ask the following questions: 1. Are the case frames in Persian and English similar? 2. If two single case frames in Persian and English are similar, are the selectional features attributed to the verb cores and the accompany- ing cases similar as well? 3. Are there any grammatical criteria common to both languages by which one can classify pro- positions and ultimately their related case frames? 4. Are the case frames in Persian translatable into English and vice-versa in a one-to-one correspondence? 1.3.1. What points of conflict exist between the two case systems of Persian and English, and what ambiguities might this cause in trans- lating one into the other at the clause level? Hypotheses The following hypotheses are advanced in answer to the questions above: 1. Preliminary observations tend to indicate that in some, though not all, examples the verb-role relationships are similar in their primary senses.2 However, how these relationships surface differentiates English from Persian, and probably all languages from one another. Selectional units in each verb determine the nature of selectional units in the nouns associated with it. In other words, as Chafe says, "... selectional units within the verb determine not only the number and relation of accompanying nouns but also, to a limited extent, the selectional units within those nouns" (1971:114). This relationship is lang- uage specific (though similar relationships in different languages are frequently found). All verbs in a language are either Stative or non-stative. Non—stative verbs, in turn, are either Process, Action or Action-process. This classification was initiated by Chafe (1971). It was used by Cook (1971-73;78) and fully adopted by Longacre (1976). As a general framework, the above classification can be applied to Persian. However, it was anticipated that not all vertical feature classifications (such as intentional, reflex- ive, etc.) advocated by Longacre (see Chapter two) for English are applicable to Persian. The feature subclassifications show diversity and deviation between the two languages. Certain features such as completability, inten- tionality and so on may be represented differently on the surface of English and Persian. In other words, the surface structures in one language are not necessarily as trans- parent as those in the other language. It was anticipated that there would be many similar case frames which are signalled differ- ently on the surface of the two languages. The differences in mapping procedures might show great conflicts, particularly when trans- lating from one into the other. 10 1.4. Data Collection The corpus upon which the study of case frames in Persian was based was selected primarily from a well- known novel -- Daijan Napeleon by Iraj Pezeshkzad (1972). Nevertheless, in cases where examples to certain case frames were not found in the original corpus, I relied on my own intuitions and those of other native speakers to create examples for the cells which remained empty in the chart. 1.4.1. Procedures In order to provide sufficient examples for the analysis of cases in Persian, all of Pezeshkzad (1972) was analyzed clause by clause, retaining the indicative clauses but excluding questions and imperatives. Similar examples were excluded, thus resulting in a preliminary corpus of 965 (out of a total 1075 possible clauses). In order to have a more limited and controllable corpus, 482 sentences representing various predica- tions were selected on a one-in-every-two random selec- tion. These clauses were first decomposed into their underlying minimal clauses, then their clause-level structure was examined. Each clausal pattern of the corpus sentences was isolated, identified, and ... \.. ll assigned a classification based both on the features of the predicates (e.g., state, process, action-process, experiential, physical, etc.) and on the arrangements of its co-occurring elements. The sentences were translated into English and edited by a Ph.D. candidate in linguistics who is a native speaker of English. In cases where the translations did not seem to accurately convey the meaning of the original sentence, they were discussed with the editor, who then attempted to provide an acceptable English equivalent. In the process of translating clauses from Persian into English, and matching the concepts in the two languages involved, examples were found whose underlying concepts either did not exist in the target language or were vague (see 5.0 in this work). Examples of this kind were excluded from the corpus. The final corpus contains 405 examples exhibiting different pre- dication identifications. To eliminate the possibility of subjective analysis and to reduce the possibility of interference based on introspection, the interpretations made on predicate- role relationships at the clause level by the researcher were subsequently judged by two native speakers of Persian and three Americans.3 Finally, presuming that Longacre's analysis of case frames and the interpretations accompanying each 12 predication presents a true picture of English, case frames in the two languages were matched to reveal the diversities which exist between them. 1.5.1. Historical Background The elaboration of case analysis -- recently developed by Fillmore (1966 ff.) -— is not a new phenomenon. Plato's distinction between 'onoma' and 'rhgma' as two constituents of a sentence reflects his under- standing of semantic correlations with 'actor' and 'action', respectively (Robins 1970:191). This binary system was later adepted and extended by Latin grammar- ians (Robins 1970:193). This development in analysis never ceased to improve. The Stoics devised further distinctions and differentiated between nominative versus oblique cases. Examples such as: l. SakratES peripatei 'Socrates walks.‘ versus 2. Sdkrétei metamélai 'Socrates regrets'. which probably make a distinction between 'actor‘ and 'experiencer' (in modern terminology) indicate the Stoics' depth of grammatical understanding (Robins 1970:195). In the era of the Greek grammarians (about the second century B.C.), the divisions made between case, tense, gender, and so on as related to the grammatical 13 categories were probably the most significant steps taken towards category distinction. Developments carried out by Dionysius Thrax and others, especially Thrax's eight class system of parts of speech, were definitely milestones (Robins 1970:189). Thus, this tendency to differentiate between the so-called 'logos' (sentence) constituents was a matter of controversy even in remote antiquity. Thrax's influence continued to dominate all linguistic research of the Middle Ages. One could still witness grammarians' efforts to establish their grammatical foundations on the traditional eight-class system. Lily's (1542) grammar of English was a significant document showing this tendency (Robins 1974:110). Some years later, Bullokar (1586) still included the five Latin classes of nouns, namely, nominative, genitive, dative (in his terminology, 'gainative'), accusative, and vocative (Baker 1931:529). Bullokar's 'gainative' is the beneficiary case in Fillmore and others' terminologies, since it is signalled by the prepositions of and for (Baker 1931: 529). Gil (1619) distinguished six cases. He included the ablative form (i.e., with+ NP) which was missing from Bullokar's grammar (Baker 1931:529). The ablative 14 case was distinguished from nominative, accusative, and vocative by the fact that the former, as with the genitive and dative forms, was marked by prepositions whereas the latter were without them. Two decades later, Butler (1633) diverted from the traditional six class system into a two class system. Butler gave the general names of oblique and rect to these two classes of nouns. Rect included the nomina- tive case as well as all other oblique cases (with or without prepositions) except the genitive case, which was considered the real oblique case (Baker 1931:529-30). During this era, one could notice a fluctuation between the two poles of formalism and functionalism. This fluctuation ranged between the recognition of six cases (Gil) and two cases (Butler) (Baker 1931:531). However, what was of greatest significance was the dominant influence of the Latin grammarians. This atti- tude towards the classification of nouns based on the Latin system was still observable in Murray's English Grammar (1795) in the eighteenth century (Robins 1974: 121-22). Thus, as seen, case analysis (in its tradi- tional form), even in the nineteenth century, had not progressed from where it had been in late sixteenth century or even much earlier than that. 15 1.5.2. Contemporary Non-transformational Scholars In this section, we do not intent to enumerate the scholars of case and predication analyses chrono- logically. The term 'non-transformational' is attributed to those case-grammarians who have developed their concepts of predication and cases without reference to generative grammar. These scholars indeed make dis— tinctions between deep and surface structures. semantic and syntactic levels, grammatical forms and grammatical meanings, and content plane and expression plane. However, in relating these levels, they never derive one form from another. 1.5.2.1. de Groot In a series of articles, de Groot (1957, 1966) discusses the significance and the problems existing in the classification and uses of cases. In his article, "Subject-predicate analysis"(l957), he tries to make a general distinction between grammatical subject and logical subject. He defines the latter as a subject of assertion (301). According to de Groot, this distinction dates back to Aristotle, who observed that "a logical subject need not be in the nominative case" (1957:301). De Groot draws distinctions between sentences and assertions by the fact that: .‘ .\4 .1 16 a. a sentence may not contain an assertion, b. a sentence may contain more than one assertion, and c. different sentences might have the same assertion (1957:304) Examples such as: 3. John died 4. It was John who died. are said to have identical assertions. This shows de Groot's insights in realizing the fact that topicali- zation does not change the meaning of predications (see Chapter Two of the present study and Longacre 286-309). As far as ambient predications are concerned (to use Longacre's termin0108Y). de Groot (1957) enters a logical discussion in analyzing clauses such as 'It's raining', 'Il pleut'. He concludes that these predi- cations “do not contain explicit reference to something (assumed to exist), but only to one, namely a process, the process of raining,..., which is affirmed to exist” (1957:304). This so-called formal subject (zero, empty, or dummy subject in modern terminology) does not refer to anything. De Groot defines a clause as a word-group consisting of a grammatical subject and a grammatical predicate l7 (1957:306). According to him, "The clause ... primarily expresses a belief in the existence of something" (307). To de Groot, intonation has a semantic function in the analysis of clauses. He distinguishes the meaning of a word from the meaning of a case. A case, according to him, "is a form-and-meaning, or phonemic-semantic, feature of a word. Consequently, what we call "the meaning of a case" is "a feature of the meaning of a word" (1966:191). Following Kurylowicz (1949), de Groot makes a distinction between primary and secondary functions of cases. The meaning of a case is fully actualized in its primary use. With regard to this distinction, de Groot gives examples of the Latin dative case, which denotes "a relation between a process and a person whose interest in the process is involved" (1966:191). The secondary function of a case, on the other hand, is the 1 use of case in syntactic-semantic 'units which contain a noun...in the given case, but which have a meaning of their own that is conventionally dependent upon certain syntagmatic, syntactic, and/or lexical feature of the unit" (1966:191). Although the major portions of de Groot's discussion concern cases in Latin, his insights in differentiating between attitudinal and referential, as well as primary versus secondary uses of cases, provide a background for later classifications of the semantic units in other 18 languages. 1.5.2.2. Halliday In a series of articles, "Notes on Transitivity and Theme” in English (1967-68), Halliday has developed the foundations of systemic grammar. In his first article (1967), Halliday's assumptions center around clause and predication, with each clause containing one predication (38). He classifies clauses into two major subcategories of extensive (i.e., clauses of action process-type), and intensive (i.e., clauses of ascrip- tion process-type) (1967:42). In extensive clauses, the predicator selects verbs of 'non-zero' types (42), whereas, in intensive clauses, the predicator selects verbs of the zero-type (47). Other features (either individually or in combination) differentiate other types of clauses with subsidiary subcategorizations. Thus, what distinguishes (5) and (6) below: 5. John cooked food. 6. The food was cooked. are the features operative, goal-transitive in (5), and receptive, agent-oriented in (6). However, both are extensive, effective clauses. Nevertheless, in the following sentence: 7. The patient looked safe and sound. 19 the subject is neither a goal, nor an actor, but an attribuant to whom an attribution is ascribed. One significance of Halliday's analysis of nominals within his system is the assumption that goal and bene- ficiary cases have more commonalities than distinctions (58). Halliday makes a distinction between goal and range.4 He claims that range and goal differ in that: a. range cannot be reflexive, whereas goal can (58) e.g. 8. *You can climb yourself. b. predicates with goals can be sibstituted by 'do', but ranges cannot (58-61). e.g. 9. Did you paint the house? No, but I'll do it tomorrow. 10. Did you make mistakes in dictation? *No, but I did in composition. Halliday (1967) also draws a distinction between depictive clauses -- those whose nominal elements are attributes characteristic of attribuants -- and resul- tative clauses in which the nominal is the result of a process (63-64). e.g. 11. He drinks his coffee black.(depictive) 12. My mother washes clothes clean.(resultative) These can be paraphrased as (l3)and (14), respectively. 20 13. He drinks coffee and when he drinks it, it is black. 14. She washes clothes, and when she washes them, they become clean (but not *they are clean). In dealing with nominals in predications, however, Halliday (1968) classifies nominals without taking into account the fact that the nature of predicates pre- determines the features in the nominals accompanying it. Consider the following examples. The case markings are those of Halliday. Page references are in parentheses. 15. John hears a noise. (181) actor goal 16. Mary turned the light on. (183) actor goal 17. Mary sat the baby up. (183) initiator actor 18. The baby sat up. (183) actor 19. The baby was sat up by Marv.(l83) actor instrument As seen, in (15), Jbbb is considered to be an actor Whereas he does not really perform any action, but is rather an experiencer whose nervous system registers the sound waves. Moreover, a noise in (15), is neither a goal nor a patient, but rather a range since it cOmpletes the predication. The 2321 in (17) and (19) is a patient whose function is different from that in (13) where it is an actor. In (19), Mbby is not semantically anything but an agent, since the topicali- zation of patient in (19) does not distort the function 21 of agent, which is not now in focus. In his article (1975), Halliday emphasizes the significance of different meaning interpretations of clauses as providing the basis of a perfect analysis. In other words, he assigns different roles to arguments as interpretations change. We will conclude this section by summarizing the example he depicts in his paper. Interpretations follow each example (345-47). 20. THE TEACHER TAUGHT THE STUDENT ENGLISH. a. actor process beneficiary goal 'The teacher imparted English to the student.’ b. actor process goal range 'The teacher instructed the student in English.‘ c. initiator process actor range 'The teacher caused the student to learn English.’ d. initiator process cognizant range 'The teacher enabled the student to come to know English.‘ e. initiator process speaker range 'The teacher enabled the student to become a speaker of English.‘ Although Halliday's analysis are not specifically case-oriented, his insights have nevertheless been very influential for case grammarians who chronologically followed him. 22 1.5.3. The Contribution of Tagmemicists 1.5.3.1. John Platt Platt's (1971) Grammatical Form and Grammatical Meaning is a tagmemic view of case concepts somewhat related to Fillmore's (1968a) deep structure and case analysis. Platt's attempt is to diverge from Fillmore's generative-semantic views and incorporate more of the traditional tagmemic concepts into case analysis. He develops a matrix based on work of his predecessors Becker (1967) and Fillmore (1968a). Platt distinguishes between grammatical forms (CF) and grammatical meaning (GM) as two fundamental dicho- tomies on which deep-surface relationships are based. He very explicitly claims that most languages have common GMs whereas GFs are more language specific (145). Notions such as subject and object are surface realizations and language-particular. To locate them on the clause level, criteria should be devised on rules of surface ability in each individual language. However, GMs are apparently universal, since any language has underlying benefactive, agent, and so on (63). Despite Fillmore's position (1968), verb or predicate is central for Platt, whereas cases or roles are peripheral or at a secondary importance (7). In presenting a clause-generating device, Platt advocates certain steps, the first of which is to select a GM 23 (i.e., of a categorical notion). The GMs thus pro- vided have to be matched to predicate fillers. As an example, he cites that if, for instance, the selected GMs happen to be a 'participative' (see below) and a 'neutral' (see below), then a verb such as 523 is most likely to occur as the predicate filler (146), since its GM implications are [+neutra1] [+participative]. Platt's grammatical meanings are as follow: a. affective: the entity affected by the action or state (73). b. agentive c. benefactive: the entity who typically benefits from the action d. factitive e. instrumental f. locative g. neutral: an entity which is in no way affected by the action or state h. participative: the entity who mentally, emotionally or sensually registers the action (Fillmore's (1970), Chafe's (1970), Cook's (1971), and Longacre's (1976) 'experiencer'). i. purposive Platt's benefactive and locative are furthermore subdivided into Inner, Outer, and Far—outer case forms (see Platt's 1971:74—78 for more details). Platt labels the doubled-up functioning of GMs as a 'portmanteau tagmeme', where two GMs and two GFs are realized by one filler constituent (148). According to him, in the following example: 21. Joe helped Tom cut the timber. Tom has an objective grammatical form in relation to 24 help and a participative role as its grammatical meaning. Furthermore, it is a subject in relation to cut and has an agentive function (148). Some of Platt's definitions of cases are vague and his examples can be distributed among other cases. Note some of Platt's examples of the participative case (59-60). The underlined cases are those under discussion: 22. Fred likes hamburgers. 23. George irritated Claude. 24. Mary pointed out the Cultural Center to Doris. 25. Bloggs murdered a warder. 26. Jack saw the film. 27. Henry knows the answer. 28. He was liked by all his staff (79). 29. Cynthia whispered to Claude (79). Here, Doris in (24) is a goal to whom an action is directed without affecting her at all. The warder in (25) is a patient who physically undergoes a change. Claude in (29) is either an experiencer if he perceives the sound waves or a goal if the action of whispering is directed to him without him being affected by the waves. In the other examples, the underlined cases are ex- periencers (Platt's 'participative'). Platt's insights on Outer-locative and Far-outer locatives are significant, though they do not seem to be within the predication and can be disregarded here. For more details, the reader is referred to Platt (1971). 25 1.5.3.2 Longacre Since the theoretical framework of this dissertation is based on Longacre's Anatomy of Speech Notions (1976), a whole chapter has been devoted to his treatment of categorical notions and predicates. The reader is referred to Chapter Two of this dissertation and to Longacre's original work for more details. 1.5.3.3. 2223 Cook has been pursuing improvements in case theory since Fillmore (1968). In a number of descriptive articles the first of which dates back to 1970, Cook follows new movements in case analysis. His own innova- tions and contributions appear in his article A Case Grammar Matrix (1972b). In this article, Cook makes efforts to provide examples for those case-frames which remained unspecified in Chafe's (1970) analysis of case frame system (see section 1.5.4.2). Chafe (1970) does not elaborate action-experiential and action-benefactive case-frames, though he sets up the framework for both cells. Cook (1972b) presumably postulates two levels of deep structure. The derived frames are said to be at a shallower level of deep structure, whereas the frames from which the derived ones originate are at a deeper level (1979:56). According to Cook, in the derivation 26 of one case frame from another, certain processes such as deleting case roles, adding case roles, and using coreferential roles are effectively put into operation (1979:57). Cook distinguishes between an unintentional object (Longacre's 'instrument') acting as a surface subject and an intentional nominal instigating an action. Consider the following examples (Cook 1979:57): 30. John (being present) frightened the baby. + {—0.13} 31. John (deliberately) fightened the baby. + [-————A,E] John in (30) is an unintentional object, whereas John in (31) is an agent. Only in example (31) can one obtain a corresponding passive. In his recent article, A Case Grammar Matrix Model (1979)5, Cook tries to incorporate and amalgamate features of Fillmore (1978-71), Chafe (1970), and Anderson (1971) into a new framework. Cook assigns one or more arguments to each predi- cate. Predicates are central elements in the semantic configuration of predications and predetermine the categorical roles occurring with them (1979:201). He arranges the arguments in a subject choice hierarchy with the highest ranking case adjacent to the verb (1979:200). 27 Cook assigns subject and object to the realm of grammatical relationships. He uses five propositional cases including agent, experiencer, benefactive, object and locative in the explanation of predicate-case rela- tionships. Time, instrument, manner, cause, result, outer locative, and outer beneficiary are considered to be optional models not required by predication (1979: 202). Cook excludes action-process verbs from his classi- fication, though his action-experiential case-frames partially cover this category (1979:203). The main objection to Cook's latest analysis is that he sacrifices a comprehensive analysis of cases in order to reduce the number of cases. For instance, Cook considers 'instrumental case' peripheral or excluded from the valence of the verb. However, we can find examples where instrument is so attributed to certain verbs that it is lexicalized within the predi- cate e.g. 32. The hunter trapped the deer. (lexicalized) 33. Johnny speared the crabs. (lexicalized) (For further discussions and details on the necessity of assigning an instrument case in the predication, see Longacre 1976:25,59,64,295). 28 1.5.4. Case grammar and transformational grammar Chomsky's Syntactic Structures (1957) marked the beginning of a new revolution in linguistic analysis. Chomsky's work redefined the goals and methods of linguistics. Chomsky questioned some of the structural- ists' assumptions by stating that the application of sets of Operations to primary data cannot lead to adequate description of grammar(s). He rather indicated that an adequate linguistic description of grammar should be viewed as a formal deductive theory. Chomsky, in his 1957 model, did not explicitly elaborate the nature of the rules which establish the relationships between syntactic structures and meaning. Meaning as it was defined by Bloomfield to include all social, cultural and individual context of speech was rejected by Chomsky (1957:99-100). Although Chomsky gave a secondary importance to meaning in the descrip- tion of syntax, he mentioned the fact that meanings of sentence constituents (morphemes) and their references are requirements to understanding a sentence (1957:103- 104). According to Jackendoff, Syntactic Structures showed that "a linguistic theory in which meaning is determined at least in part by a level of underlying structure can capture important generalizations” (l972:l). Chomsky's quotation "I think that we are forced to conclude that grammar is autonomous and independent 29 of meaning..." (1957:17), at least to me, shows his major concern in 1957 with the formal syntactic devices of the language though he never disregards the role of meaning. In his words, "Despite the undeniable interest and importance of semantic and statistical studies of language, they appear to have no direct relevance to the problem of determining or characteri- zing the set of grammatical utterances" (1957:17). Newmeyer also defends Chomsky's 1957 position and claims that "... he [i.e., Chomsky] regarded as theoretically significant a whole set of systematic connections between syntax and semantics" (1980:32). Chomsky's Aspects of the Theory of Syntax (1965) ‘began new controversies over meaning in general and the mechanism relating deep structure to surface Structure in particular. For Chomsky in 1965 the deep Structure was the input to the interpretive semantic Component. Chomsky's phrase structure rules were responsible for providing functional and categorical information. However, the main concern was to define the "derivative relations between categories" (Newmeyer 1980:87). For his new recent thinking, see Chomsky 1975, 1978, and 1979. 30 1.5.4.1. Fillmore's Case Grammar Newmeyer (1980) in his discussion of the emergence of case grammar refers to the position of generative grammar in the late 19605. In his words "the relatively shallow deep structures of Aspects were attacked from another quarter" (128). Fillmore developed his case grammar in the context of transformational grammar. Fillmore's case grammar, however, is not essentially different from general transformational linguistic theory, though it differs from the Aspects model in significant ways. Fillmore's Case for Case (1968) makes certain claims, the most salient one being its claim of the applicability of the theory at a language universal level. Chomsky (1965) and Fillmore (1968) share some basic assumptions on formal and substantive universals. B0th Chomsky and Fillmore place universal constraints on the base component forms. Thus, they tend to develop the base component in such a way so as to Provide comparable terms to all individual languages. According to Chomsky (1965), "A theory of substantive universals claims that items of a particular kind in any language must be drawn from a fixed class of items" (28). And furtherfore, he claims that "The existence of deep-seated formal univerals ... implies 31 that all languages are cut to the same pattern, but does not imply that there is any point by point correspondence between particular languages"(l965:30). What worries Fillmore with respect to the Aspects model is, as it seems to him, an inability of that model to handle the problem of the representation of both functional and categorical information conveyed in the deep structure of phrases (Newmeyer 1980:128). Fillmore's interpretation of the Aspects model of deep structure apparently makes him take the position that it cannot capture the double function of phrases such as by me, in the bag, with his fist, etc., which are both prepositional phrases and agent, location and instrument, respectively (Newmeyer 128-129). It seems that Fillmore's main objective is to add as much semantic information as possible to the deep Structure. Fillmore argues that the notions such as "Subject of..." or "object of..." lack the required Semantic value to describe associations within the Sentence. In a series of articles (1966-71), he deVelops the case relationships and includes notional Categories such as agent, experiencer, counteragent, and so forth in his grammar. His main concern, as mentioned above, is to close the gap between semantic representations and the deep structure. Fillmore sets up a small number of case relations 32 so that each clause contains the maximum amount of deep structure information. Thus, a sentence node domi- nates a modality and a proposition which, in turn, includes a verb form and a number of cases associated with it. Fillmore claims that case relationships exist univer- sally in all languages. In Fillmore's theory, the term 'case' refers to the semantic-syntactic relationships which exist under- lyingly in the deep structure of clauses. According to Fillmore, the case relationships are not restricted to cases reflected by affixes; they can also be expressed by grammatical devices such as word order, clitic particles and so on. Thus, Fillmore deviates from the more traditional subject-predicate categorization made by transformationalists. He assigns a much deeper semantic interpretation to deep structure than could be found in Chomsky's Aspects model. As Fillmore states, "...the subject/predicate division is an importation into linguistic theory from formal logic of a concept which is not supported by the facts of language...", and furthermore, "...the division actually obscures the many structural parallels between 'subjects' and 'objects'" (1968a:17). Here, I will briefly critique Fillmore's (1968a) lIlodel: 33 A. Advantages a. Case features (such as source, performer, and so on) are not inherent characteristics of nouns. The features previously attributed to lexical items by genera- tive grammarians such as [common], [abstract], etc. were more absolute and permanent properties of the nouns. b. Fillmore (1968) assigns a hierarchy to the eight cases he develops in his theory. Thus, there is a high predictability of which case has the potential to be promoted to subject position. c. Fillmore's theory reveals the similarities between verb frames such as those of bbby and £33» which are different on the surface (Newmeyer 1980:129-131). B. Drawbacks a. The equative clauses (Pike and Pike 1977) -- the NP is NP construction type -- are missing from Fillmore's (1968a) case grammar. He probably disregards them since their deep structure configurations are different from those in which an action is involved. b. Fillmore's 'objective' case is a waste-basket which absorbs any noun whose function cannot be subsumed in other case categories. Thus, the objective case ciefined by Fillmore as "the most neutral case" is a dumping place for all unspecified cases. To demonstrate lh<3w Fillmore's objective case can further be analyzed as other cases, a set of his examples will be presented. 34 All the underlined nominals are in Fillmore's objective case. The cases in parentheses on the right are my analyses of the cases (roles). Figures are page number references to Fillmore (1968a). (This list has been adopted from Platt 1971:26). 34. Joe removed the book. (27) (Patient) 35. Joe gave Mary the book. (27) (Patient) 36. Joe opened the door. f (27) (Patient) 37. Joe saw the film. (30) (range) 38. Joe showed Fred the book. (30) (range) 39. Joe likes chutney. (30) (goal) 40. Joe knows Fred. (31) (goal) 41. Joe looks at the film. (31) (goal) 42. Joe learned the lesson. (31) (range) 43. Joe heard the sound. (31) (range) 44. Joe listened to the music. (31) (goal) Platt(l97l:27) shows that Fillmore's own test of "affectum-effectum" cannot be applied to the predications containing see, please, like, know, look, and so on, since the so-called 'objectives' in these predications are not affected by the actions or states. Thus, what Joe did to the book was remove it is acceptable, but what Joe did to the film was see it is unacceptable (Platt 1971:26-27). 1.5.4.2. Chafe Chafe's Meaning and the Structure of Language (1970) Presents substantial modifications to the theory of <2ase analysis. His generative semantic views6 provide Szlgnificant incorporations to the theory of language in gieneral and to the noun-verb relations in particular. 35 Chomsky's disparity between deep and surface structure has influenced Chafe's (1970) framework, though Chafe never considers deep structure anything but semantic structure (9). The innovation which makes his work substantially different from that of Chomsky(l965) and Fillmore(l968a) is the assignment of a greater role to verbs with regard to other accompanying nouns. According to Chafe, a semantically well-formed structure requires a major semantic element, namely, a verb, whose function is to narrow the conceptual field. Thus, a verb occupies the central position as the core in a semantic structure(1970:10;346). According to Chafe, the elements in the semantic structure are not lineraly ordered(44) though there is a hierarchy among noun-verb relations(349). Chafe postulates a set of postsemantic processes-- including linearization, deletion, and literalization -- whose main functions are to transform semantic structures into surface structures(l970:29;54). These processes involve reshuffling semantic configurations to provide the linearized backbone of surface patterns. This deep- to-surface switching in not a sharp sliding process, lDut rather a gradual process which starts with the Slnrinkage of semantic features and agglutination of plionetic features(70). 1 36 According to Chafe, the whole "human conceptual universe" is primarily bifurcated into two major areas of l) verbs, and 2) nouns. Verbs refer to actions, event, qualities, and so forth, whereas nouns are more peripheral, centering around verbs(96). Despite Chomsky(l965) and Fillmore(l968a)7, Chafe assigns the greatest role to verbs. Thus, it is subcategorical features of verbs which predetermine the cases on nouns. Chafe postulates a set of testing criteria in order to classify all verbs into the four major subdivisions of state, process, action-process, and action(l970: 98-99). Process verbs are accompanied with patients but not agents. Action verbs require at least an agent whereas action-process verbs demand both patient and agent (1970:100). Chafe's horizontal parameters consist of basic verb types; experiential; benefactive; and locative. Chafe offers no case frames for action- experiential and action-benefactive types. Some case frames are basic, while others are said to be derived by the application of certain rules. For instance, what distinguishes £22 from Ebby, and learn from teach is an optional causative rule which applies on the former of each pair to derive the latter(Chafe 1970: 3128-9; 146 and Cook 1979:58). Chafe's system includes seven cases. These cases aloe agent, patient, experiencer, beneficiary, instru- InGent, complement, and locative (1970:100-156). 37 Here, we will try to enumerate Chafe's contribu- tions to case theory. These will be followed by the mention of drawbacks. (All references are to his 1970 work). A. Contributions a. Chafe assumes that verbs are core elements of predicates and predetermine the functional roles co- existing with them(96). b. He proposes a universal classification of verbs into State, Process,Action-process, and Action (98-107). c. A distinction is made between ambient case frames versus non-ambient case frames, with ambient referring to clauses with a single predicate and no nominal in the semantic pattern of which the predication is made. For instance, in It's hot, the meaning of the sentence involves nothing but a predication (101). d. Chafe hypothesizes a feature 'potent' in the semantic specification of a noun which can be super- ficially categorized as agents, but which do not possess the required characteristics of agents. In other words, it is an extension of the agency of animate entities to inanimate entities which may occur as agents. e.g. 45. The heat melted the butter. where bbbb is potent, but not a true agent. e. A deviation from a pure syntactic analysis to a more semantico-syntactic perspective is proposed. 38 With regards to meaning, Chafe takes the position that "meaning is what language is all about, and a linguis- tic description must reflect this simple fact" (Langacker 1972:149). f. Chafe distinguishes between examples such as (46) and (47): 46. Tom is hot. (experiencer) 47. 122 is hot. (patient) In (46), Tom's nervous system registers the environmen- tal condition whereas, in (47), 222 is hot as someone else experiences it (147). g. Chafe disagrees with the generative semanticist notion that constituents at the semantic level are in a linear order.8 To Chafe, ordering of constituents at the level of semantic representation is quite improper (Langacker 1972:153). B. Drawbacks a. Chafe does not distinguish ambient state predications with no nominal at the surface level from ambient experiential state predications with an animate entity experiencing the environmental conditions. Consider the following example: 48. 1:2 freezing here. where l is an experiencer whose nervous system undergoes the environmental conditions. b. Chafe's conception of patient is very broad. 39 Consider the following examples. Page references are in parentheses: 49. Tom wanted a drink. (144) 50. Tom knew the answer. (144) 51. Harriet broke the dish. (103) The noun-verb relations in (49) and (50) are different from those of (51). In (51), the dish undergoes a physical change, whereas, a drink and the answer are apt to be complements rather than patients. Harriet in (51) is an agent who instigates an action. This is different from the situation in (49) and (50), where 123 is an experiencer. c. Many examples which are provided to illustrate the beneficiary and locative cases can be incorporated into one single 'goal' case. Note the following examples. Page references and Chafe's assignment of cases are presented in the parentheses. 52. _EE has the tickets. (beneficiary, 147) 53. om found the tickets. (beneficiary, 149) 54. Mary bought 122 a convertible. (beneficiary, 149) 55. Tom threw the knife into the box. (locative, 161) 56. Tom fell off the chair. (locative, 160) Flt-l In sentences (52), (53), and (54), 22$ is the entity in which the predications terminate. The same is true in (55), where the box is the point in which the action terminates. In (56), the chair, despite Chafe's claim that it is in a locative case, is a source since it is the locale which the predication assumes as place of origin.9 40 d. In order to indicate how grammatical classes switch on the surface, Chafe posits certain deep structure units and adds to them a suffix :32 to derive other grammatical units. Thus, widen (a process verb) is derived from 3135 (a state) by the addition of :33, Some other verbs go through opposite processes. Thus, broken (state) is derived from break (a process verb). There are still other verbs such as bpbb, which undergo no derivation in examples such as: 57. The door is open. 58. John opened the door. 59. The door opened. Longacre (1976:234) shows that the grammar will be less haphazard if we postulate a feature called state in the deep structure of EASE and broken, but a feature called process in the deep structure of widen, and bpbb. e. Chafe's definition of instrument is misleading. Instrument -- as Chafe defines it -- is "subsidiary to the agent, [that is], something which the agent uses" (152). According to him, in a sentence such as: 60. The key opened the door. one can infer an agent which uses the instrument to instigate an action. Thus, we can have a parallel sentence: 61. Tom opened the door with the key. However, in the following sentence: 62. The wind opened the door. 41 no parallel sentence comparable to the example (61) can be found. We cannot say: 63. *Ali opened the door with the wind. f. Chafe does not distinguish between complement and measure as two different deep cases. Consider the following sentences: 64. Mary sang a song. (Chafe's complement: 156) 65. The book weighs a pound.(Chafe's complement: 157) However, these two so-called complements are semantically different deep cases. Longacre has shown that they are different. We can have (66) as corresponding to (64) but not (67) as correspondence to (65). 66. A song was sung (by Mary). 67. *A pound was weighed by the book. g. Chafe does not seem to differentiate between peripheral and non-peripheral locatives. Consider the following examples: 68. The knife is in the box. (159) 69. Mary danced under a tree. (162) It is true that both sentences answer a non-polar question starting with where, however, they are sub- stantially different. In example (68), in the box is a non-peripheral (i.e., nuclear) locative and is obligatorily present as part of the predication, whereas in (69), under the tree can be Optionally deleted. 42 1.5.4.3. Anderson(197l) Case, as defined by Anderson, refers to grammatical relations. These relationships determine the nature of noun participation in sentences which are charac- terized by verb features such as process or state (10- 11). Anderson attempts to demonstrate how a relation- ship can be established between the concrete and the more abstract uses of certain roles which surface identically (5). Anderson's scheme for relating the underlying case relationships to surface realizational configura- tions has two components: 1) a semantico-syntactic component and a morphologico-phonological component; and 2) a lexicon. A set of subcategorization rules is involved in developing complexes of categories. On the other hand, a set of constituency rules expands symbols into their constituents. Both of these sets of rules apply at the semantico-syntactic level (19-20). In Anderson's system, verbs and adjectives are categorically identical. They are characterized with respect to the features [t stative] (37-38). Moreover, Anderson distinguishes ergative clauses from non- ergative clauses. The surfaceability of subjects determines the clause ergativity as in sentences (70) and (71) below: (Anderson's examples 40 and 41) 70. Egbert read the book. 71. Egbert killed the duckling. 43 According to Anderson, both subjects (but not the objects) in 69 and 70 represent underlying ergatives. Moreover, both 70 and 71 can have progressive and imperative correspondences. However, as in (72) and (73), the superficial subjects do not correspond with the underlying ergatives: 72. Egbert knew the truth. 73. This bag contained the money. The fact that (70) and (71) can have corresponding imperative and progressive sentences whereas (72) and (73) cannot differentiates these two categories (40). Thus, what differentiates the verbs such as 5&1; and die lies in the fact that kill is attributed [+ergative] whereas g33_is [-ergative]. Anderson concludes that big and bib; are inherently the same (44-45). Anderson's major contributions to case theory and his innovations are summarized below; this is followed by the discussion of certain drawbacks in his localist view of case. A. Contributions 3. Anderson draws a distinction between reflexive versus non-reflexive clauses, as in (74) (reflexive) and (75)(non-reflexive): 74. Egbert moved. 75. Egbert moved the couch. (49) b. He differentiates between two types of transi- tive clauses as in (76) and (77): 44 76. Egbert damaged the book. 77. Egbert read the book. Anderson (64) argues that these two are different by virtue of the fact that a question such as What did Egbert do to the book? can have a response as it is in (78) but not (79). 78. He damaged it. 79. *He read it. However, one could argue that the surface subjects in (76) and (77) are two different categorical notions at the deep structure level. Egbert in (76) is an agent/initiator whereas, Egbert in (77) is an experiencer/ agent. c. Anderson acknowledges the fact that certain surface intransitive verbs have underlying covert objects which rarely surface. In other words, these so-called intransitive clauses are more transitive- like underlyingly but intransitive superficially. Consider Anderson's examples (67): 80. He drinks. 81. Egbert is painting. 82. They kissed. which can be extended to: 80a. He drinks beer/wine. 81a. Egbert is painting a painting. 82a. They kissed each other. d. Anderson makes a distinction between locatives in which the noun within the locale indicates the spatial location of the nominative and those locatives which are not intrinsically part of the nominative 45 specification. Anderson's examples are as follows: 83. He remained in London. 84. We keep the money in the box. According to Anderson, these locatives are marked on the surface of English sentences by the case markers 13 or 33 (81)- e. Anderson draws a distinction between: 85. John is cold. (referring to John's sensation) and 86. John is cold. (referring to someone else's sensation upon touching)- To be more explicit, he distinguishes between experien- tial clauses as in (85), and attributive clauses as in (86) (96). B. Drawbacks a. Anderson apparently does not consider the double function of nouns on the clause level. Consider the following sentences. Page references are in parentheses. 87. Egbert read the book. (40) 88. Egbert killed the duckling. (40) 89. l have a compas with/on me. (113) 90. Egbert sneezed. (175) 91. The ball rolled across the floor. In example (87), Egbert is both agent and experiencer. Egbert in the example (88) is an agent. 1 in (89) can be considered either as a goal if it is interpreted as the permanent possessor of the object, or a path if it is interpreted as a transitional owner. Egbert in 46 (90) is a patient. The ball, in the example (91) is a patient. b. Anderson does not distinguish between 'goal' and 'locative' on the one hand, and 'source' and 'locative' on the other hand. Consider the following example: 92. The ball rolled from Jane to Mary. (119) In this example, Jane is a source whereas Mary is a goal, since the action terminates at her. Anderson considers both of these roles/cases as locative cases. 1.5.5. Case Grammar and Stratificational Theory To the best of my knowledge, very few works have been published which extensively discuss stratificational views on case-predicate relationships. Lockwood (1972) and Sullivan (1980)10 are the only scholars whose works are familiar to me who treat the problem of identifi- cation of participants at the clause level to any great degree. Since Lockwood's is the most widely-available model, we will limit the discussion to his work. In stratificational theory, language is held to consist of relationships. A set of connections associate conceptual relations and phonic correlations, both of which are outside of the language (Lockwood l972:6). In addition to certain realizational relations which connect it to higher or lower strata, each stratum has . . ll . its own syntax or tactics. In general, during encoding, 47 higher strata control lower strata and the immediately lower stratum in particular. In stratificational theory, sememic structures identify participants in predications, predications generally being the realizates of clauses at the lexemic level, which roughly corresponds to surface syntax. In Lockwood's model, the structure of the clause is detailed in the lexotactics. However, the semotactics is responsible for the well-formedness of propositions (172). Thus, the sememes -- the tactic units of the sememic stratum -- represent roles and predicates in the sememic configurations. In Lockwood's model, elements in sememic structures are unordered, though the lexotactics partly linearizes the realizations of the sememic events and their accompanying participants. The lexotactics produces combinations of constituents which may not be sensical by themselves; this is allowable, however,since it is the function of the semotactics, not the lexotactics, to characterize the set of well-formed propositions. The basic structure of a predication (the sememic realizate of a clause) consists of an event sememe (e.g. the realizate of a verb) accompanied by one or more roles (cases) in a case-frame. Lockwood apparently views agent, goal, recipient, instrument, causer, and beneficiary as major roles, whereas time, location and 48 manner are "circumstantial attributes to the predication" (1972:142). Lockwood's use of agent and goal are not clear. Note the following examples. Page references are in parentheses. The case identifications are those of Lockwood's. 93. Myron drapped the rock. (142 agent) 94. Penelope saw Sammy eat anchovies. (163 agent) 95. My aunt died. (167 agent) However, an analysis of verb categories indicates that neither Myron, nor Penelope nor my aunt can be real agents, since none of them instigate an action. The first sentence might, in my judgment, have two inter- pretations as follows: 93a. Myron dropped the rock on purpose 93b. The rock fell off Myron's (hands) acciden- tally If (a) is understood, then Myron is an agent, otherwise he is a source from whom the action originates. Both (94) and (95) above are experiencers since Penelope's and the aunt's nervous systems register the actions, respectively. They do not answer 'what did Penelope/ your aunt do?’ but rather can occur as acceptable answers to 'what happened to Penelope/your aunt?‘ Lockwood does not present examples to show the distinc- tions between causer and agent. Lockwood does not distinguish between a locative case (an obligatory case in a case-frame which may be 49 covert, partially covert or totally covert on the surface) and a 'place', which has a peripheral role to the predication. Lockwood, however, implicitly excludes the latter as a major case based on the fact that it cannot be focused. As mentioned earlier, sememes are arranged hier- archically rather than linearly. Although Lockwood implicitly gives a greater role to verbs in the selection of major accompanying roles (1972:143), his sememic configurations do not explicitly indicate the centrality of verbs as core elements of predications. It is not clear whether verb subfeatures predetermine the nature of accompanying nouns or vice-versa.13 Lockwood's (1972) model of linguistic relationships indicates that stratificational theory has great potential as a framework for describing case-predicate relation- ships. (For more details, see Lockwood 1972:136-189). 1.5.6. Case Analysis and Functional Grammar (Dik) Tagmemic theory regards language as primarily a means of communication, and its followers generally believe that the analysis of language should not be restricted to sentences, that language is best analyzed as one aspect of human behavior, and that language has more than a symbolic function. In 1977, Pike says 50 .. language is not abstracted from life, but is merely one part Of it, Operating on principles necessary for all purposeful actions" (xvi). This is in line with his previously-expressed views. Dik (1979) has once more raised the issue of language as the most significant instrument operated by human beings in social interaction. According to Dik, a language analyst should seek to set up two kinds of rules: 1) pragmatic rules, to govern "verbal inter- action as a form of cooperative activity", and 2) semantic, syntactic, and phonological rules as instruments on this activity (1-2). The analysis of a language should principally and practically reach pragmatic adequacy, psychological adequacy, and typological adequacy (6-8).14 In functional grammar, there are three different levels of language, each specified with certain functional relations: semantic functions -- including notional categories such as agent, goal, recipient, etc.--syntactic functions, related to surface manifestations of roles, and pragmatic functions including tOpic and focus, and theme and tail (Dik 1979:13; and 1980:49-50). Roles or arguments refer to "entities in some world" (1980: 51), whereas predicates establish coherent relationships between these linguistic primitives (Dik 1979:15-16). According to Dik, predicates which are not formed 51 according to synchronically productive processes are basic, while Others are derived predicate frames (1980:52). The analysis Of linguistic expressions begins by assigning to each linguistic construction an underly- ing predication. Rules (not in a transformational sense) then Operate to map the underlying representations onto those on the surface to determine the surface order Of constituents. The underlying predication, in turn, consists of terms -- cases or roles -- incorpora- ted into predicate frames (Dik 1980:51). On the other hand, the lexicon consists of sets of terms and predi- cates containing unpredictable and underivable features attributed to them. In other words, a predicate is specified with prOperties which indicate the category to which it belongs, the lexical form it has, the number of categorical notions permissible in its domain, the semantic functions each participant has, and the selectional restrictions imposed on the accompanying roles (cases) (Dik 1979:30-31; and 1980:52). Two fundamental parameters, namely dynamism and control, specify the state of affairs (states and processes). Any action involving a change potentially possesses dynamism: such actions are events. On the other hand, if any of the arguments determines the occurrence or non-occurrence of the state of affairs, 52 the predicate is basically controlled. The amalgama- tion of pluses, and minuses of the features [dynamism] and [control] leads to four types of states of affairs: actions, positions, processes, and states, the last of which is [-dynamism, -control] (Dik 1979:32-35). Thus, in 96. John smashed my bicycle. 97. John stayed at home. lgflfl is said to be the controller who controls the state of affairs. Dik makes a distinction between nuclear arguments within a predication and satellite arguments, which are peripheral to the predication. What governs the peripheral or the satellite arguments is the nature of the state of affairs rather than the inherent features of predicates themselves (1979:50-51). Source, locative, direction, manner, instrument, time, duration, cause, purpose, reason, etc. are involved in the specification of the semantic functions of satellites (1979:50). Dik assigns a semantic function hierarchy to arguments; this hierarchy is based on the potential of arguments to be promoted to subject or object positions. Thus, agent, goal, recipient, beneficiary, instrument, locative and time are hierarchically ordered to such a way that agent has more potential to occur in the function of subject. The potentiality for subject/object promotion is directly related to the position of an argument in the hierarchy (1980:63). See Dik 1979 and 1980. Notes to Chapter One See Chapter Two for more details. 2I will differentiate between the primary senses of verb-role relationships and what we can call the 'configurative' relationships. In the latter, many factors such as cultural and religious backgrounds determine the relationship. For example, the sentence: 1. The sea claimed many lives (Platt's example: 1971:20) and its Persian equivalent: 2. darya qorbEni-ha-ye besyar gereft sea victim-pl - E many took-it ldarya/ (sea), in both languages is considered pri- marily more like the agent than the instrument. However,it seems to me that in a configurative relation- ship, the speaker of Persian considers it as an instrument used by God, who is then the real agent. 3 . . My intention was to distribute questionnaires concerning the different interpretations for each clause to native speakers of Persian. A preliminary attempt showed that this would not succeed because of social factors. Thus, the sentences in the study were presented orally to two native speakers of Persian who agreed to give their interpretations. In addition, besides the editor, two Americans were consulted. l'Longacre adOpted Halliday's 'range', see Chapter Two for details. 5Cook's Case Grammar (1979) is a collection of his previous articles. 6According to Langacker, "Chafe's theory is Offered quite explicitly as an alternative to transformational grammar, but it is more Of an alternative to some versions of transformational grammar than to others. Much of his discussion is directed against the standard 53 54 theory and its successor, interpretive semantics..." (1972:146-47). Furthermore, Langacker claims that, "If his [Chafe's] theory resembles case grammar and generative semantics in certain respects, it is in part because he anticipated some of the innovations embodied in these theories." (1972:147). 7In Fillmore's system, nouns are nuclear and thus predetermine verbs, which are peripheral (Fillmore 1971:38). 8What differentiates between syntacticism and seman- ticism is that syntacticism postulates a level of deep structure between the surface structure and semantic representation. Chafe identifies deep structure with semantic structure (1970:9). Chafe as well as generative semanticists rejects the notion that a deep structure stands intermediate to surface and semantic structures (Langacker 1972:149). However, Chafe's views and the views of generative semanticists do not always match (Langacker 1972:148). According to Langacker, "Chafe has expressed disagreement with two specific aspects of the semantic structures proposed by generative semanticists. First, he claims that the linear ordering of constituents is imprOper at the level of semantic representation ... Second, he doubts that the embedding of logical pre- dictions will allow one 'to account for the various meaningful relations between semantic nouns and verbs'..." (1972:153). 9For more details, see Chapter Two of this work and also Longacre 1976:23-37. 10David Bennett has implicitly referred to cases at some considerable length in his published dissertation Spatial and temporal uses of English prepositions (1970). However, he has not tried to universalize them. 11The tactics in stratificational theory are as follows: gnostotatics, semotactics, lexotactics, morphotactics, and phonotactics. For the function of each tactic, see Lockwood 1972: chapters 4, 5, 6, 7, and 8. According to Lockwood, agent in his system of description of English is surface-oriented. In other words, it refers to whatever occurs unmarkedly as subject. However, he admits that agent and causer might be different gnostemically (personal communica- tion, 1981). 55 13Lockwood states that, to him, classes of nouns and verbs are simply said to concur in their given roles. According to him, whether verb subfeatures predetermine the nature of accompanying nouns or vice-versa is a pseudo-question (personal communication). 4According to Dik, a grammar of language should achieve pragmatic adequacy, psychological adequacy, and typological adequacy. a. pragmatic adequacy A grammar should reveal ...those properties of linguistic expressions which are relevant to the manner in which they are used, and to do this in such a way that they can be related to a description of the rules governing verbal inter- action"(Dik 1979:6). b. psychological adegpacy "A grammar should ... be psychologically adequate in that it should not be incompatible with strongly validated psychological hypotheses about language processing: (Dik l979:7). c. typological adequacy According to Dik, a theory of language should be capable of providing different grammars for typologically different languages. It should also be capable of demonstrating similarities and differences which exist between these languages (1979:8). CHAPTER TWO Theoretical Framework Two significant areas of linguistic theory serve as the foundation of this analysis of the case-frame system in Persian: 1) Longacre's case-system analysis Of English developed in his book An Anatomy of Speech Notions (1976); 2) tagmemic theory as developed in general by Kenneth Pike in his monumental work Language in Relation to a Unified Theory of the Structure of Human Behavior (1967) and, in particular, in Pike and Pike's recent work Grammatical Analysis (1977). 2.1. Longacre All grammatical analyses of the case-systems of English and other languages which have been done so far have attempted to indicate -- with as much precision and completeness as possible -- the roles of constitu- ents and the ways in which they relate to each other in a sentence. Longacre's system is not only a further step in this direction, but also a significant contribution towards a more complete analysis Of case frame systems in English. Longacre's analysis (1976) is an investigation of notional categories which, he claims, underlie 56 57 all languages. His analysis is an attempt (along with his predecessors: Fillmore(l968), Chafe(l970), Cook (1971-75), Platt(l971), Anderson(l97l) and so on) to analyze the universal or universal-like categories of case. According to Longacre, the distinction between grammatical categories (i.e., subject, object, etc.) and situational categories (i.e., actor, goal, etc.) is an old tradition, but was never the focus of attention until the late sixties (1976:23). The fact that surface structure categories represent highly sophisticated functional categories and also the fact that there is a roughly one-to-one correspondence between the functional categories and the linguistic primitives of the real world provide a general basis for Longacre's theoretical foundation(1976:23). Longacre sorts out the roles underlying surface forms in such a way that each role gets its own term which represents its function more clearly. Consider the following examples, based on some of Longacre's illustrations: John patted pg on the shoulder. John knifed the attacker. John cut the log. John broke his leg. ran a race. built a house. caught sight of Mary. John listened to music. John loves his wife. (... O 'J‘ :1 (.4 O 3‘ :3 \OCDNO‘U'IJ-‘UJNH C... O :3‘ 5 58 lbbb, as the surface subject of the clauses, has different underlying roles. In (1), lbbb does something to someone who experiences the effects of the action. In example (2), lflhfl does something to someone who physically receives the action. The attacker is wounded by a special kind of instrument applied by prp, who intentionally uses it. In example (3), gppp is an agent. In example (4), John's inalienate possession physically undergoes a change. In (5), gbbb_is an agent, whereas, the race is simply a range and does not undergo any action. In (6), £222 performs an action, as a result of which something previously non- existent comes into existence. In (7), Jbbb uninten- tionally is the one who perceives or experiences the light beams. In (8), as in (7), £232 is an experiencer, with the only difference being that he intentionally attends to the sound waves. In (9), not only does John's nervous system register the emotion, but the emotion also has a certain direction towards a goal. On the other hand, the so-called surface 'objects' in these sentences are underlyingly 'experiencer' in (1); 'patient' in (2), (3), and (4); 'range' in (5), (6), (7), and (8), and 'goal' in (9). Now, if we expand the sentences above in such a way that each clause gets further possible cases, more varieties of 59 surface structures will result (1976:24-26). The following examples are taken into consideration: 10. John patted me on the shoulder with his hand. 11. John cut the log with an axe. 12. The axe cut the log. 13. The log was cut with the axe. In example (10), with his hand is an unmarked instance of the instrumental case in English since the verb 'pat' necessarily implies the application of hands as means. In (12), 223 has preempted the subject position at the surface level without changing its underlying functional role. Similarly, in (13), lbg is still a patient no matter which surface position it occupies. Longacre concludes that relationships between the constituents are not restricted to surface configurations, but also obtain in the underlying patterns(1976:26). 2.1.1. Predication According to Longacre, a complex of features within the case frame represents the predication in the deep structure. The predication, in turn, is represented by the verb and accompanying nouns on the surface. In other words, features such as State, Process, Action and Action-process describe predications in the deep structure. At the same time, predications correspond underlyingly to the verbs. The features may or may not be accompanied by certain characteristics, namely, 60 physical, motion, locative, completable, intention, reflexives, posture, measure and instrument (1976:19). Like Fillmore's case theory (1968), Longacre's analysis of predication stipulates the nonlinearity of predi- cates and their related nouns in accompanying roles. In other words, a case-frame includes a predicate with one or more associated nouns in a non-linear order (1976:19;306-307). 2.1.2. Cases or Roles Like Fillmore, Longacre considers structural relationships to be of two types: a. surface structure relationships b. underlying relationships The underlying relationships do not have fixed surface representations and, within the surface structure patterns, "... the structure is linearly ordered" (1976:307). As Longacre states, "It [surface structure] is not a mere end product of the process of transformational derivation. Rather, choice of one surface structure against another is a meaningful option on the part of the speaker" (1976:296). Focus and tOpicalization are two options for the speaker's selection, among others. These surface structure phenomena may high- light subjects, objects, instruments and so forth. 61 Topicalization, however, cannot change the functions and the roles each participant has in the underlying information (1976:27). The inventory of cases that Longacre provides as the backbone to his analysis of the case-system in English is essentially built on his predecessors' contributions to case analysis. Anderson's (1971) localistic view of case relationships has been parti- cularly influential in Longacre's (1976) work. As Longacre himself admits, Anderson's influence prompts him to assign greater roles to goal, source, and path (1976:27). All other cases except measure, have already been used by Fillmore (1968), Platt (1971), Cook (1971), Chafe (1970), and Halliday (1967) as summarized by Longacre (1976:25). Measure is a case which has not been used before except by Longacre (1973) (See 2.1.2.5. in present work). A general comparison between Longacre's (1973) case inventory and his most recent analysis (1976) shows the changes which have taken place in his approach. For instance, source, path and goal have been promoted from subcate- gories of locatives (1973) to different independent categories in his more recent work. In the analysis of Persian case frames, I have used other terms which I borrowed from Pike and Pike (1977), and Sullivan's (1980). However, these notions 62 were incorporated into Longacre's framework.2 Here, in order to make it easier for the reader to understand Longacre's theoretical foundation, which will be thoroughly applied in this work with only minor modifi- cations, we will discuss the 'cases' and illustrate them with simple examples. (For the remainder of this chapter all references to Longacre will be to his 1976 study, unless otherwise indicated). 2.1.2.1) Experiencer: Longacre refers to nouns as experiencer in the following situations: a) one who reacts to his environment, e.g. 14. l feel cold in this room. b) one who is characterized by an emotional state, e.g. 15. K2 are glad the hostages are freed c) an animate entity whose psychological state is disturbed in one way or another, e.g. 16. John was brain-washed before he was released. d) an animate entity who is impinged upon by someone's activity, e.g. 17. John scared pg to death by hiding behind the door. d) an animate entity which experiences violence, affection, etc. e.g. 18. Mary patted pg on the shoulder. f) an animate entity which has acquired knowledge, e.g. 19. Michael knows English, French, German and Swedish. 63 g) one who receives sound or light waves, e.g. 20. I heard him curse. 21. b s aw him come. h) an animate entity whose nervous system registers love, appreciation, hatred, etc. e.g. 22. The people of the world love one another. In all of these examples, the action impinges on someone's nervous system. (27-28). 2.1.2.2) Patient a) an animate or inanimate entity which undergoes a change of state or location, e.g. 23. He removed the glasses from the table. 24. He pushed pg toward the door. b) an animate or inanimate entity that changes state or location with or without an agent affecting it through an action, e.g. 25. The horseman fell from the horse. c) an inanimate entity which is the prOperty of someone, e.g. 26. President Reagan has a nice £3223: 2.1.2.3) Agent An animate entity which acts or instigates an action or an inanimate entity which runs, revolves or functions, e.g. 27. Ali killed the duck. 28. The factory is still running. 2.1.2.4) Range Any nominal that specifies the predicate or is the end result of that predicate, e.g. 29. The Iranian troops have fought a winning battle against the Iraqis. 64 30. Beethoven composed several beautiful symphonies. (29-30) .2.l.2.5) Measure "...the surface structure nominal which completes a predication by quantifying it..." (Longacre 1976:30). e.g. . 31. I have lost twenty-five pounds since my wife left. 2.1.2.6) Instrument a) an inanimate Object intentionally used by an instigator to carry out an action or a process. e.g. 32. I'll shoot him right in the heart with this gun. b) a covert instrument built in to the predicate.3 The covert instrument is redundant in English. e.g. 33. We finally trapped that mouse. c) an animate entity acting as a stimulus unin- tentionally. This instrument emotionally triggers the state of mind. e.g. 34. Some people are scared of mice. Based on Longacre's definitions, one is apt to believe that a nominal might function either as an instigator or an instrument. What differentiates the two is whether the predicate implies intentionality or uninten- tionality. Compare (35) and (36) below: 35. John is scared of his mother-in-law. 36. John is often scared by his mother-in-law. 65 In sentence (35), John's mother-in-law is an uninten- tional instrument. In sentence (36), John‘s mother- in-law is the one who occasionally terrifies John with her behavior. (31-32;54-57). 2.1.2.7) Locative According to Longacre, source, goal and path (see 2.1.2.8, 9, and 10 below) often replace the locative case, which is more restricted in use. Longacre refers to a locative as the locale where some action occurs, without implying any motion or direction through a place. Consider the following sentence: 37. The Statue of Liberty stands 12 New York. There is no movement to, from or toward New York (32). 2.1.2.8) Source Source as a case refers to a nominal in the following functions: a) the locale which serves as a predication's origin. e.g. 38. Margaret Thatcher left her luxury hotel and went to the White House. According to Longacre (32), the nature of the verb determines the function of the nominals accompanying it. Thus, a predicate containing verbs such as 'throw' requires a source coreferential with the agent, e.g. 39. A i threw the ball over the fence. 66 Here, Ali is the agent as well as the source from which the propulsion begins. b) "...the entity from which physical wave relevant to sensation emanates..." (Long- acre 32). Consider the following example: 40. Kamran smelled the roses and bought a bouquet. where 'the roses' is the entity from which the smell originates. c) the animate entity which gives away physical possession. e.g. 41. Zari sold all her property before she left East Lansing. where Zari is the agent and the original source or owner of the property. 2.1.2.9) Goal The goal is: a) the entity at which a predication terminates. e.g. 42. All demonstrators were led to the pavilion. b) any animate entity which is the recipient of a predication without being physically changed. e.g. 43. Michael loves his cat. Here, love is directed towards the cat without the animal being physically touched upon by the action. c) the non-transitory owner, whether a new or a permanent owner. e.g. 44. Dr. Jones has a beautiful house in East Lansing. 45. Steve was awarded a $500 tax-free grant. 67 2.1.2.10) Path The path is defined as: a) the locale an action traverses through with- out that locale or locales being the source or goal of the predication. e.g. 46. The railway crosses hundreds of miles of desert. b) a temporary owner of property. e.g. 47. On the way home, 1 purchased a gold bracelet with diamonds on it for my wife. In (47), b is the 'agent/path', whereas EiEE is the 'goal' and also is/will be the non-transitory owner. According to Longacre, 2232.15 the only case, in English, which can theoretically reoccur in a single predication without restrictions. (34-35). 2.2. Non-nuclear cases Longacre agrees with Cook (1972a:46) that certain cases (those listed in 2.1.2.1 to 2.1.2.10 above) are the most essential cases within a predication. They are essential because their occurrence in the predica- tion is 'diagnostic'. In other words, they are necessarily required as components of predications. On the other hand, other cases are peripheral and not essential to a particular verb. According to Longacre (1976:35-37), these cases are as follows: 2.2.1 Peripheral locative4 Peripheral locative refers to the palce where an action occurs without it being a necessary component Of 68 the predication. Thus, according to Longacre (35), in a case analysis one should distinguish between a non-peripheral versus a peripheral locative. Consider the following examples: 48. Masqasem put some glasses on the table close to the plates of fruit. (Nuclear locative) 49. In the dim light of the courtyard, I was overcome with fear. (Non- nuclear/Peripheral locative) 2.2.2. Time, MannerL Cause, Purpose, Accompaniment Longacre considers time, manner, cause, purpose, and accompaniment (Fillmore's 'comitative',5 l9683:81) as peripheral cases whose occurrence in the predica- tion is not required on the clause level. These, according to Longacre, are pertinent on the sentence 1evel(35). 2.3. Parameters In order to categorize the verbs of English based on diagnostic features, Longacre (1976) devises a set of vertical parameters which are characterized by certain values. Longacre adOpts Chafe's (1970) verti- cal parameters, namely, State, Process, Action-process and Action as valid classifiers of verbs(1976:43). In other words, Chafe's distinctions between State and Nonstate; Process and Non-process verbs (Chafe 1970:98-104) have been adopted and fully applied by Longacre. 69 As far as horizontal parameters are concerned, Longacre distinguishes three general categories using the three features: Ambient, Experiential, and Physical. Each category has its own subclasses, which are characterized by subcategory features/values. Thus, he proposes a chart containing 45 filled cells with three empty cells. The subcategories have something in common semantically (50). .In my judgment -- and as Longacre himself admits (50) -- Longacre's ordering of values are not neatly organized, in contrast with his predecessors; however, he gives more details about the subcategories. In order to provide a brief but precise perspective of Longacre's parameters, a set of examples follow (for Abbreviations see P. viii). Some of these examples have different analyses in Persian. See Chapter Three for the Persian case-frame categories. (Longacre 38-97). fS-Amb 1 It's rainy today. P-Amb 2. It's warming up. A AP-Amb 3 God caused it to warm up. (Longacre 1976:44) A-Amb 4. It's raining now. \ ( S-Amb -Exper 5. I'm hot. P-Amb -Exper 6. I'm getting too hot here. B AP-Amb -Exper 7. I cooled myself off with ' the air conditioner. \ A-Amb -Exper 8. I got caught in the rain. I S-Exper 9. My cousin is angry. P-Exper 10. My cousin got mad after C he heard the news. AP-Exper 11. My cousin's wife made him angry. A-Exper 12. Romeo patted Juliet on \ the shoulder. S-Exper -Complet C' P-Exper -Comp1et AP-Exper -Complet A-Exper -Comp1et \ ' S-Exper -Directed P-Exper -Directed D AP-Exper -Directed \ A-Exper -Directed ’ P-Exper -Directed -Comp1et D' AP-Exper -Directed -Complet A-Exper -Directed ‘ -Complet ’ S-Phys P-Phys E AP-Phys LA' S-Phys -Measure P-Phys -Measure F AP-Phys -Measure A-Measure S-Phys -Loc AP-Phys -Loc A-Phys -Loc 70 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. I know Stratificational theory is a new theory. I forgot your room number. Dr. Jones teaches history. John is studying math. Men love their families. I fell in love with a beautiful woman. John introduced his wife to me. I'll sue your company. I saw a dragon with my own eyes. We told the children stories before they went to bed. My son listens to pOp music everyday. This book is interesting. My father died when I was 15. The guerrilas killed the captives. That fat man eats a lot. I weigh over 180 lbs. now. I have lost a lot recently. President Reagan will cut taxes by 10% this year. I won ten dollars playing poker. My car is in the garage. Masqasem put the glasses on the table. A 12-month old baby should be able to stand on its feet. G! H! 71 (P-Phys -Motion 35. The pencil rolled off the table. AP-Phys -Motion 36. The police pushed the demonstrators towards the lobby. [ A-Phys -Motion 37. Johnny swam in the river for two hours. S-Phys -Poss -Directed 38. I had a car last year. P-Phys -Poss -Directed 39. I got my M.A. in 1974. AP-Phys -Poss -Directed 40. Michael sold me a ticket to the Opera. A-Poss -Directed 41. I keep all my mother's letters. ’ S-Phys Poss -Motion 42. I have candy for each child. P-Phys Poss -Motion 43. My brother finally found a new place to live. AP-Phys-Poss -Motion 44. I bought my wife a new gold bracelet. ‘ A-Phys-Poss -Motion 45. Mrs. Behers collected stamps for the needy last year. As seen in the above examples, what differentiates between A and B; C to D'; and E to H' parameters are the features ambient, experiential and physical, respectively. On the other hand, what differentiates A and B is a subfeature experiential. In B, there is an animate entity whose nervous system registers the environmental activity. Parameters C and C' are differentiated by the fact that C' has an extra sub-feature completable. In other words, the nominals known, learned, taught or studied are ranges in relation to animate entity whose nervous system registers a state or an action. An optional instrument feature might co-occur with the 72 examples in Parameter C. D and D' differ from C and C' in that the former contain an extra feature directed. The predicates in these categorical predications indicate directionality Of action. However, in D, the existance Of a goal necessitates the direction of action towards a goal or a termination of action in the goal. E and F differ in that an obligatory case measure occurs in F but not in E. As mentioned earlier, measure is Longacre's innovation in case analysis. The features locative versus motion and possession-directed versus possession-motion differentiate G and G' and H and H', respectively. For the details of each parameter and examples of further derivations, the reader is referred to Long- acre (1976). 2.4. Deep and Surface Structure According to Longacre, the deep structure Of a clause is more transparent than the surface structure in showing the case relationships. In other words, functional roles such as agent, source, patient, etc. are more clearly seen in deep structure than in surface structure. That is to say, functional roles are more meaningful at the deep structure level. However, Longacre assigns meanings to constituents such as subject and object, though their meanings are more 73 opaque in contrast with the functional roles. Subject as part of Old information and object as the new information have their meaning at the surface level. (287). Longacre does not separate deep structure from grammar. In other words, according to him, notional or deep categories are an essential part of the grammar itself. (11). In his words, ”...deep and surface structure seem to be very similar in broad outline so that the roles and case frames which we find in predications are not unlike the tagmemes and clause types of surface structure. The former are simply a more consistent set of categories than the latter" (288). Longacre's argument for including deep structure within the grammatical hierarchy is based on the fact that, to him, the deep -surface relationship is not a dichotomy but rather a distinction of two poles. As he states, "...a certain distinction should not be assigned thoughtlessly either to the deep or to the surface structure but that we have to evaluate its relative depth" (289) (emphases is mine).6 Longacre further classifies surface structures into two poles: elastic versus non-elastic structure. Surface elasticity refers to the degree of surface expansion over a variety of deep structures (288). A clear example is transitivity which, on the surface, 74 represents different predication categories. On the other hand, surface structures such as ambient predi- cations or, according to Longacre, more peripheral predications such as equatives or 'descriptives' are less elastic (288). Longacre's classification of surface structures on the relative axis transparent versus opaque is a deli- cate distinction. The transparent surface structures are relatively less marked and clearly reveal their deep structure configurations. On the other hand, the opaque surface structures are heavily marked and are "more autonomous of the deep structure" (291)- According to Longacre, "... nuclear tagmemes (see below) are less transparent...by comparison with peri- pheral tagmemes. The deep structure shines through the Peripheral tagmemes..." (294). Longacre's major concern in this issue is whether any change of meaning should be attributed to a necessary elaboration of deep structure. In other words, he wonders whether the choice among similar surface structures is a meaningful one. First, Longacre takes the position that no exact paraphrase exists and that any rewording on the surface will necessarily lead to a change of meaning(294). Second, he rejects Chafe's (1970) position that surface structure is post-semantic (for further details see Chapter One Section 1.5.4.2.)- 75 To Longacre, despite Chafe(l970), Fillmore (1968) and Anderson(l97l), surface structure is not the result of a transformational operation. Where two or more surface structures are available, the speaker's selection of one versus another is a meaningful choice (295-96). 1 will conclude this section with an illustrative quote from Longacre: "Surface structure, while pri- marily the domain of form has some meaning correlates. Likewise, deep structure, while primarily the domain of meaning, has some formal correlates. Here as elsewhere every linguistic unit is a form-meaning composite"(301). 2.5. Tagmemics as a model In the analysis of Persian proposed in this work (Chapter Three), I use a tagmemic model for the grammatical descriptions wherever required and expedient. The separability of levels in tagmemic theory provides a unique opportunity to bring single levels into focus Of attention. Moreover, the nature of the tagmeme as it has been developed by Pike and Pike (1977) represents the combination of class, function, and filler. Grammatical form, grammatical meaning, and the filler all appear in a single unit, and are formulized on a single line. This helps to illustrate the nature of 76 deep-surface relationships more clearly. The emphasis by tagmemics on language as an in- separable portion of human behavior makes it one of the best candidates for the description of language as used in communication. As Adam Makkai states, "Pike's system...offers an organic, self-contained and mature theory of language in relation to human behavior, ... it includes a serious attempt at classifying cultural institutions of all sorts. In this regard, Pike has gone farther than any of his contem- poraries. If any of the competing schools of thought in contemporary linguistics intends to do justice to language as a cultural institution and not just its disiecta membra, to quote Hjelmslev, all the facts and aspects of human cultural behavior discussed by Pike will have to be incorporated into the frame of refer- ence"(l972:43-45). The essential characteristic of tagmemics is its concept of hierarchy. In other words, small units are constituents of larger units, and this hierarchy expands upward towards larger constituents or units. Languages, in general, have different hierarchical levels, each of which may include units on the same level or on different levels. The different hierarchies are not dichotomous and are simultaneously interrelated (Jones 1980:79-80). A tagmeme is a constituent with four features represented in four cells; there is a close relationship between these four features in such a way that both 77 meaning and form are involved in this complex. A tagmeme and its features can be illustrated as follows: slot class role I cohesion 'Slot' refers to the function which is manifested by a grammatical form and which represents the grammatical relation of the whole unit to other units on the same syntagmatic axis. The 'class' or 'filler' is the entity which fills the slot. "Role' encodes the deeper functional or semantic relationship of the entity which occupies the slot. "Cohesion' is the cell where typo- logical relationships between different constituents are encoded. These cohesive relationships differ at various levels. They might be agreement features at the clause level, sequence time at the paragraph level or story time at the discourse 1evel(Pike and Pike 1977:35-43). An example might clarify these points. In a sentence such as Ali loves his wife, Ali can be represented by a four cell tagmeme as follows: Subject I + undergoer (experiencer) # > On the clause level, the slot is the subject filled by a proper noun, namely, Ali which represents an under- goer at the deep level. The cohesion cell specifies that the subject tagmeme governs number agreement in the predicate, thus, Ali loves..., but not Ali love.... 78 The undergoer, in turn, is represented, at a deeper level, by a semantic case 'Experiencer'. In principle, Longacre emphasizes the significance of tagmemics for grammatical analysis at various levels (For details, see Longacre 1976:255-309 and Pike and Pike 1977:35-54). Tagmemic theory is the most apprOpriate descriptive tool for analyzing the case-frame system of a language for the following reasons: 1) As mentioned above, tagmemics has the potential to focus on constituents at different levels separately or at all levels simultaneously. In other words, the notion of hierarchy is integral to a tagmemic description of all langu- ages (Jones 1980:78). 2) Tagmemics does not make one level more prominent than another. Tagmemics is a system of behavior. Thus, morpheme is not separable from clause and,' in turn, a clause is encompassed in monologue discourse and dialogue exchange (Jones 1980:85). The tendency to consider units within larger units may be very helpful in the analysis of case-frames in all languages, particularly Persian. Consider the following Persian example: (Pezeskzad 1972:13) 79 50. ba bil zad—am tuye qag-e dahan-es with shovel hit-I in split-E mouth-his/its 'I hit him in the teeth with a shovel'. A general tagmemic formula for example (51) might be as follows: NP 5 ro n.> Marg. + S

f Marg. N + Rel. ' position ' Nuc I NRt + Marg. Psr.Suf. Item I Psr. I At this level, the existence of an obligatory Margin as possessor filled by a possessor clitic implies that the patient has an obligatory feature of animateness. Inanimateness is not marked. For instance, we have sentences such as these: 51. ba samgir zad-am tuye sang with sword hit-I in stone 'I struck the stone with a sword.‘ 52. b5 lagad zad-am tuye be sandali with kick hit-I 'in/to chair 'I kicked the chair.‘ NOTES TO CHAPTER TWO 1This is not the case in Persian, in which the act of patting is not restricted to physical action; it may be verbal as well. Thus, the following sentence is acceptable in Persian: 14. zan-et-o b5 kalam-e xos navazeg kon wife ~your —Ach with word-E nice patting do (lit. Do pat your wife with nice words.) 'Be nice to your wife.’ 2Cook (1971) has used 'place' for peripheral locatives versus 'locatives' for nuclear cases (Longacre 1976: 35). Cook's terminology will be used in the descrip- tion of case-frames in Persian to illustrate this distinction. Certain instrument cases are obligatorily covert in English. Thus, we never say I ate food with my mouth, teeth, etc., except where more specifications are required. For example, in a sentence such as The man kicked the chair with his left foot, ‘foot' as a tool has been specified by the attributive 'left'. Nuclear cases or roles are necessary constituents of predications. Nuclear roles are part of the valence of the verb, whereas peripheral roles are optional entities. At the predication level, we use the terms 'nuclear' and 'essential' interchangably. Similarly, 'peripheral' and 'non-essential' both refer to the Optional entities which are not relevant to the specification of underlying predication. 5Fillmore makes a distinction between the following two sentences and mentions the possibility of a comitative case. According to him, fig and his wife have a conjunction relationship in sentence (1), whereas, his wife in (2) is a comitative case. 81 82 1. He and his wife are coming. 2. He is coming with his wife. Sentence (1) may have the following different inter- pretations: a. He is coming from the right and his wife is coming from the left. b. He is coming from the right and his wife is also coming from the right but separately. c. He and his wife are coming from one direction and together. Now, if we replace the nominal within the so-called comitative case with some other nouns, ambiguities might arise as the result. Consider the following examples. 3. The mother is coming with her baby. 4. The mother is coming with her daughter. In sentence (3), the baby is not in the comitative case but is a patient being carried by her/his mother. In (4), both interpretations are possible. 6Cases are parts of case-frames; case-frames underlie clauses on the surface and predications at a deeper level. Based on Longacre's analysis (305-306), the following simple chart might be devised to show the deep-surface relationships at different levels in one language system. Deep_Levels Surface Levels Plot Discourse Repartee Paragraph Proposition. Sentence Predication Clause Concretion Phrase Inflection Word Derivation Stem 83 However, these deep-surface level relationships do not have a universal one-to-one correspondence. These relationships are rule-organized, and the mapping processes must be spelled out by rules (Longacre:305). CHAPTER THREE Case Analysis of Persian 3.1. Introduction In this chapter, I present a detailed analysis of the cases and case frames of Persian. This chapter includes nine major divisions, each of which includes a number of subheadings. The major divisions are l) introduction; 2) devices for case identification; 3) nuclear or diagnostic cases (i.e., those essential to the predication) in Persian; 4) devices for predication categorization; 5)case frames (predications) in Persian; 6) passive versus active construc- tions; 7) derived case frames, including reflexive case frames; 8) mapping of deep and surface realizations of clauses including tagmemic formulas; and 9) conclusion. The analytical problems concerning case frames in general are discussed in the introduction. The theoretical assumptions, the language under study, and a brief review of the source of the data are also included there. In addition, the basic word order in Persian and the major verbal construction -- simple versus compound verbs -- and auxiliary verbs are presented in that section. In section 3.2., the nominal features--properties assigned to entities in juxtaposition with predicates-- are presented. In diagnosing cases and assigning roles to nominals, features are essential, in the sense that noun features must match those features inherent in predicates. 84 85 Nuclear cases of Persian are discussed in section 3.3; they are twelve diagnostic cases, as Opposed to peripheral cases -- non-essential elements in predication identification. Cases which show partial overlapping are grouped in single categories with subheadings to differentiate them. This categorization leads to three single-item groups and three groups consisting of more than one case. The definition for each case is followed by case markers which indicate the occurrence of cases in case frame configurations and which act as surface representations of them. In that section, I do not add any new cases, employing only those which have been proposed by other scholars. (However, the particular cases prOposed here for Persian are different from those included in earlier studies). Devices for predication categorization in Persian are presented in section 3.4. I define there the features which are essential in predication categorizations. Features such as physical, experiential, possession, etc. are included. A number of verbs are shown to be differentiated according to the accompanying features. In section 3.5, predications are categorized according to their predicates-- core elements of case frames carrying meaning-- and the nuclear cases adjacent to them. In this section, thirteen different groups of predications in Persian are presented. In each group, typical clauses which exemplify the case frames are provided. The groups are 86 as follows: A: ambient; E: ambient experiential; g: experiential; D: experiential directed involving verbs of desire and cognition; 2;: experiential completable (in- cluding verbs of sensation and speech); E and E: physical and physical measurable case frames, respectively; G: locative; EL: physical motion verbs; and, finally H and Hi: possession directed and possession motion case frames, respectively. Equative predications are included in Group I, although the nature of equatives as predications has been questioned by Longacre, whose theory of case underlies the present work. Predications are categorized according to the features state, process, action—process, and action. An integrated chart including the conflated case frames is presented. Section 3.6 presents examples which demonstrate the restrictions Persian imposes on the selection of passive constructions. The passive construction, as an alternative to surface realization of predications, is shown to be less acceptable in Persian. Longacre's criteria for the acceptability of superficial passive clauses are used to demonstrate that the so-called "passive' clauses are unaccept- able or odd with most case frames in Persian. Certain case frames with coreferential surface subjects and objects, and predicates characterized by the feature reflexive, have surface representations which include 87 reflexive pronouns. In section 3.7, we identify the re- flexive case frames which relate to the major case frames in section 3.5. How the surface and deep structures of the Persian indicative clauses are matched is the subject of section 3.8. Tagmemic formulas are provided to represent deep-surface relationships at the clause level. Transitive, intransi— tive, ditransitive, equative, and descriptive clauses are matched with their underlying predications. This chapter terminates with a short conclusion. 3.1.1. Definition of Terms The following are definitions for some of the common terms as they are used in the context of this study: Cases or Roles In Longacre's terminology, case refers to nouns in their accompanying roles. In other words, case specifies the underlying roles of noun phrases as they basically appear in surface grammatical relationships. Surface nominal entities encode cases (nuclear or peripheral) which function at a deep level. Cases may or may not be signalled by case markers on the surface. Here, we distinguish between surface grammatical categories, such as subject, object, complement, etc., and cases as deep categories, such as agent, goal, patient, range (see 3.3) etc. Note that in traditional grammar, cases are basically realized as surface 88 forms, whereas in case grammar the term 'case' refers to the semantic roles underlying surface forms. case frames A case frame refers to a verb (or a verb phrase) in juxtaposition with one or more cases (i.e., noun(s) in their accompanying roles). In other words, a case frame is defined as a semantic configuration consisting of a verb with certain characteristics, accompanied by a number of permissible roles (cases). Verbs are first classified according to whether they are states, processes, action- processes, or actions. Second, verbs are classified with respect to the semantic features attributed to them. Predicates and their obligatory or optional features are placed in brackets; roles or cases accompanying them are outside of the brackets. Thus a case frame such as {Egggfgir A E R should be read as an action-process experi- ential completable case frame with accompanying nominals in the role of agent, experiencer, and range. No linear order should be inferred from the form of the case frame itself. features What distinguishes a set of verbs from another is a number of features which characterize the verbs in case frames. Predicate features determine the allocation of verb forms in case frames at the predication level. Moreover, prEdicate features determine the nature of features within 89 accompanying cases, (i.e., predicates project features onto nominals accompanying them. For instance, the verb ko§tan 'to kill' requires an agent with a feature/intent/assigned to it when the verb itself possesses a feature/intentionall). predication Predications are notional structures which underlie the surface structure of clauses. Superficial verb forms are substituted for in the deep structure by a predicate. Predicates in turn are represented in case frames by features such as state, process, action-process, and action, and a number of diagnostic characteristics such as completable, intention, etc. (for more details, see Chapter Two; and Chapter Three, sections 3.4 and 3.5). vertical parameter Verb forms represent either state or non—state predi- cates. Non-state predicates are either process, action, or action-process. These four primary features characterize predicates in general. Thus, in a vertical parameter, verb types are classed as state, process, action-process, and action. horizontal parameter Features such as experiential, physical, completable, directed, possession, and so on characterize predicates within different case frames and help categorize case frames into various groups along a horizontal parameter. In the Present study, the groups A-I stand on the horizontal dimension. 90 function In tagmemic analysis, function refers to the purpose or role carried by a unit on the syntagmatic axis. Each slot performs a function (or role) (Pike 1977: 35; 485). margin In tagmemic grammar, margin (or periphery), in contrast to the nucleus, refers to a construction which is dependent. In a sentence such as the following: 1. ali diruz jelo-ye cesm-e hame-ye ma bedehi-ye xod-ra pardaxt Ali yesterday front—E eye-E all-E we debt-E self-Ach paid. 'Ali paid for his debt yesterday in front of us all'. jelo-ye cesm-e hame 'in front of us all' and diruz 'yesterday' are margins. The distinction between margin and nucleus is not restricted to sentences, but it may also be relevant at the phrase level. Thus in the phrase: 2. mard-an-e bozorg—e tarix man-pl-E great-E history 'The great men in history' mard 'man' is the nucleus of the phrase whereas the other constituents are marginal. £388 marker In Persian, case is usually marked by prepositions. The specific occurrence of a case marker is partially pre- dictable based on the case frame in which it occurs. However, £1 complete generalization is not quite possible. Prepositions at times can be absent from the surface but their insertion 91 does not distort the semantic relationships. Note the following: 3. ali ketab-am dad 'Ali gave me (a) book' Ali book-me gave-he 4. ali be man ketab dad 'Ali gave me (a) book' to I In these examples, be man and -am are goals. case domination By case domination, we refer to case frames in which the existence of one particular case requires the existence of another case in the same simple predication in certain environments. For instance, the occurrence of goal as a benefactive requires that there be an agent/source in the action-process case frame. The covertness of the agent on the surface does not change the relationship. Another instance would be the instrumental case, whose presence depends on the existence of an agent to make use of it intentionally (see section 3.3 for further details). The following examples should clarify this point: 5. dast-i be sane-ye man zad/ ba dast be sane-ye man zad hand-a to shoulder-E I hit/with hand to shoulder-E I hit 'He patted me on the shoulder' 6. masqasem ...rise-ye nastaran-e bozorg-e daijan-ra b5 Masqasem root-E jonquil-E big—E Daijan—Ach with kolang qat' karde bud pickaxe cut done was 'Masqasem had cut the root of Daijan's big jonquil with a pickaxe'. 92 7. man in hendi-ra be dast-e xod—am mikos-am I this Indian-Ach to hand-E self-my kill-I 'I'll kill this Indian with my own hands' 8. ingilisi-ha u-ra koste bud-and English-pl he-Ach killed were—they 'The English peOple/troops had killed him' The occurrence of b3 dast 'with hand' in (5), ba kolang 'with pickaxe' in (6), be dast 'with hand/by' in (7) require the existence of u'he [covert], Masqasem, and man, 'I' in (5), (6) and (7) respectively. The instrumental case in (8) is covert on the surface but present in the underlying predication. covert, partially covert, and overt cases Cook mentions deep structure constituents which may or may not surface (1979:82). According to Cook, overt cases refer to propositional roles implied by the predicate and always present on the surface. A covert case, on the other hand, is a case which is obligatorily present in the deep structure but never present on the surface. Finally, partially covert cases are those cases which may be absent or present on the surface depending upon the environment in which they occur. Longacre also refers to certain roles demanded by the predicate which do not usually occur in the surface structure but which might surface if certain speci— fications are attributed with them (see Chapter Two, of the present work). In the following example: 93 9. azizosaltane sohar-as-ra gaz gerefte bud Azizosaltane husband-her-Ach bite taken was-she 'Azizosaltane had bitten her husband' a covert instrumental case occurs in the deep semantic configuration. However, the instrumental case may occur on the surface in a specified extended related clause: 10. azizosaltane sohar-as-ra b5 dandEn-e gorEz-es gaz gereft Azizosaltane husband-her-Ach with tooth—E boar—her bite took 'Azizosaltane bit her husband with her sharp long teeth' 3.1.2 Data The data used in the analysis of Persian case frames are a number of clauses selected at random from a Persian novel (for more details, See Chapter 1, Sections 4 and 4.1 in the present work). As the following chart of verb entry distribution matrix shows, some cells relating to different case frames indicate no verb entries. However, in order to have an analysis including all possible case frames in Persian, examples based on my own intuition as a native speaker and checked by other native speakers are included to fill the gaps in the data. 94 Predicate Verb Action- Classes Entries State Process Process Action 382 65 136 94 87 A l 0 O O O B 0 0 0 0 0 C 51 7 35 7 2 C' 10 8 O 1 l D 14 8 3 1 2 D' 61 O 37 20 4 E 125 7 44 27 47 F 4 3 l 0 O G 23 13 0 4 6 G' 63 O 12 28 23 H 22 14 2 6 0 H' 2 O O 0 2 I 6 4 2 O O Verb Distributions Matrix For the purpose of simplicity, the following modifica— tions are made: 1) in some of the examples containing long names, common names are occasionally substituted for them, 2) in examples where nominals may have various successive adjectives attributed to them, the non-essential elements are eliminated. 3.1.3. The Problems of Case Analysis There are a number of problems in analyzing cases and aSSigning meanings to them. Some of these problems have 95 already been solved by pioneers in case analysis; others are still to be solved. Still more problems have been interpre- ted differently by different scholars. Certain questions have always been in focus. Are case frames interpreted in the same way by different members of the same speech community? Are features assigned to different roles and predicates evaluated identically by different speakers of the same language? Are utterances understood equally well if they are used in all contexts? In other words, in an example such as: 11. hasan xuk-ra ko§c 'Hassan killed the pig' Hassan pig-Ach killed-he is Hassan always an agent who intentionally and voluntarily initiates and performs the action? But what, if Hassan happens to fall off a roof and landing on the pig kills him on the spot, unintentionally? To what extent can one inter— pretation in a limited corpus be justifiably generalized to other contexts? Most case-grammarians, except Cook (l972)1, have intentionally tried to limit their data to domains where clear-cut criteria can be devised to differentiate semanti- cally well-defined dichotomies. However, authors of literary texts generally employ fewer clauses whose semantic functions are transparent, that is, elaborate forms are used to convey simple notions. It seems to me that at least in Persian, the most Opaque2 the grammatical-semantic relation- ships are, the more formal the texts. 96 Idiomatic and metaphoric expressions are two major concerns here. It is possible to divide the surface con- stituents of idioms into semantically related constituents on the basis of one-to-one deep-surface correspondence? In other words, in a sentence such as: 12. qamar del be aspirin dade bud Ghamar heart to Aspiran given was-she 'Ghamar had fallen in love with Aspiran' and its paraphrases: 13. qamar aseq-e aspiran sode bud Ghamar lover-E Aspiran become was-she 14. qamar be esq-e aspiran asir sode bud Ghamar to love-E Aspiran captive become was-she is there a unique sememic configuration which is realized differently in the surface structure? Is it possible to devise a case frame for sentence (12) which involves an action-process possession in such a way to indicate that, in fact, Ghamar is an agent/source who deliberately bestows her heart on Aspiran, who then functions as a goal? If not, can it perhaps be related to a process experiential directed case frame in which Ghamar is an experiencer whose nervous system registers the act of love which is, in turn, directed towards Aspiran, the goal? Is the expression del be kas dadan 'to love someone' a unit, or is it a sequence of units del+be kas+ dadan, with regard to its case function? These and many other questions are a hindrance to a quali- tative analysis of cases. 97 3.1.4. Theoretical Assumptions The following premises constitute the backbone of this study with regard to case analysis in Persian: a) the predication is the smallest semantic unit which carries information. A predication consists of a predicate (filled by a verb or a verb phrase) as its core element and one or more cases (roles) accompanying it. b) unlike the predicate, which is the chief meaning carrier within the predication, cases (roles) are either obligatory or peripheral. Their occurrences are dependent on the features of the predicate fillers. The peripheral categories may or may not be present in the semantic configurations of predications. c) obligatory or peripheral are not inherent features of the cases, rather predicate features determine the nature of roles that the deep cases play. Thus, in a clause such as: 15. ali barre-r5 zebh katd 'Ali butchered the lamb' Ali lamb-Ach slaughter did-he the verb zebh kardan 'slaughter' requires that an agent be present in the semantic representation of the case frame to initiate the action intentionally; it also requires an instrument for the agent to use, whether or not the instru- ment appears on the surface. In addition, the use of the verb zebh kardan instead of kostan 'to kill' eliminates any ..-.. ... ‘ . . .: . I ... e .. s a a .\ .. a u i .5. . ... . e A w. «z 1‘ .3 . . ... .» ... ... t. s a. 4. w~ -. ..e a. .3 5. . - . ,. ... ..— e ‘ g . . s .: e . . A ... a. .4. a. u» d. ..- .s .: \t. ‘- a." 98 possible interpretation that the action took place acciden- tally. d) different predicates may surface identically in different horizontal parameters. Thus, one verb form may occur with different roles in different case frames. e) Noun or noun phrases represent at least one role (case) and up to three simultaneous roles at most. Consider the following examples: l6. masqasem david 'Masqasem ran' 17. masqasem baraye baEEe-ha qese mi-goft Masqasem for child-pl story said-he 'Masqasem used to tell children stories' 18. masqasem hamise az xodas ta'rif mi-kard Masqasem always from himself description did-he 'Masqasem always praised himself' Masqasem is an agent in (16); both agent and source in (17); and an agent, an experiencer, and a goal in (18). f) nominals are neutral out of context and are only attributed with certain features in collocations with pre- dicates. Thus, at the deepest level, there are certain semantic elements which are neutral as to role and meaning. For instance, in a clause such as: u sar-e mar; he sang sekast he head-E me with rock broke—he 'He cracked my skull with rock' the semantic representation consists of three neutral nominals adjacent to the predicate. At a shallower level, 99 u, sare man, and sang take on their grammatical meanings. That is to say, 2 gets its functional role as an agent, sar as a patient, and sang as an instrument. The markers are surface elaborations 3.1.5. The language under study Persian belongs to the Iranian branch of the Indo- European family of languages. That branch includes a number of genetically-related languages which still exist and are spoken over a large territory in Asia—-in Iran, Afghanistan, Tajikestan, India, and Pakistan. It sounds reasonable to claim that approximately seventy million peOple speak Iranian languages in these areas. In Iran, Modern Persian is the language of communication and education as well as the language used by mass media. 3.1.6. Word order Persian is primarily an SOV language. However, different word orders can occur which are semantically marked and demonstrate stylistic variations. Nominal elements may also occur in different positions to express topicalization, and the so-called passivization. Consider the following examples: 19. man yekbar ezdaha-ra dide bud-am (nonspecific/ 4 specific/definite) I once dragon-Ach seen was-I 'I had seen a/the dragon once' (unmarked) 20. ezdaha-ra man yekbar dide bud-am (specific/definite) 'I had seen the dragon once' (marked) 100 In example (20), egdaha occurs clause-initially to make it more focused in the possible interpretation. However, example (20)can become more marked if a cleft construction (i.e., in +NP+budan+ke...) or a pseudo-cleft construction (i.e., anEe+...+budan+NP) is used as in the following examples: 21. in ezdaha bud ke man did-am this dragon was that I saw-I 'It was the dragon that I saw' 22. in man bud-am ke ezdaha-ra did-am this I was-I that dragon-Ach saw-I "It was I who saw the dragon' 23. anée man did-am e§daha bud what I saw-I dragon was 'What I saw was a dragon' In Persian, verbs agree in number with agent, experiencer, patient, and so on, that is, with the nominal that fills the surface subject slot. Some verbs morphologically also agree with the nominal elements being topicalized. However, the subject-verb agreement is not relevant to a) impersonal predications, and b) predications in which pronoun subject has lost its semantic valence. Consider the following examples: 24. man pa—ha-yam tab-e tahamol-e badan-am-ra na-dast I foot-pl-my power-E bearing-E body-my-Ach neg-had 25. *man pa-ha-yam tab-e tahamol-e badan-am-ra na-dast-am I foot-pl-my power-E bearing~E body-my-Ach neg-had-I 101 26. man tab-e tahamol-e badan-am-ra na-dast-am I power-E bearing-E body-my-Ach neg- had-I 27. man az in dande be an dande mi-sod—am I from this rib to that rib became-I "I was tossing and turning' 28. az in dande be an dande-am mi-kard-and from this rib to that rib—me prog-did—they 'I was tossed and turned' 29. man xater-am saxt masqul bud I mind-my hard busy was-it 'I was terribly busy' 30. * man xater-am saxt masqul bud-am I mind-my hard busy was-I In (26) and (27), the verbs agree with the surface subjects. Examples (25) and (30) are unacceptable since the verbs agree with subjects which are semantically void. The predicate in (28) requires an agent, though it is an unknown third person plural. Thus the verb form there agrees with an unknown subject. 3.1.7. Persian Verbs Verbs are used here as formal manifestations of underlying predicates. Predicates are the core elements in the predication; they carry and lend meaning to the non- predicate satellite arguments. 3.1.7.1. Simple and Compound Verbs Verbs in Persian have traditionally been divided into simple and compound types. The number of simple verbs is 102 relatively limited. In modern Persian, there is a great tendency to use compound verbs (i.e., prefix+verbal root and nominal+auxiliary)rather than simple verbs. Examples such as aqaz kardan (lit.beginning do) for agazidan 'to begin', and talab kardan (lit. debt do) for talabidan 'to ask for/ to request' are a few among many. However, in some cases where the use of one single form leads to ambiguities, both have been preserved. Consider the following examples: 31. man parid-am (ambiguous) 'I flew/I jumped' I jumped-I 32. man parvaz kard—am 'I flew' I flight did—I 33. parande parvaz kard 'The bird flew' bird flight did 34. man az ruye divar‘parid-am '1 jumped over the wall' I from on ,wall jumped-I 35. *man az ruye divar parvaz kard-am5 I Traditionally, simple verbs have been considered as those verbal phrases which contain only a verb root. Predi- cates preceded by a prefix, or a nominal followed by an auxiliary (budan'be'; kardan'do'; sodan'become'; dastan'have') are regarded as compound verbs. However, following Tabaian (1979), I believe that no justifiable syntactic and semantic criteria can differentiate between simple verbs versus compound verbs in general. Tabaian argues that 1) though the prefixes had a semantic valence in an earlier stage of the development of language, they have lost their semantic 103 Significance in the contemporary language, since the prefixes are no longer productive (196; also Windfuhr 1979: 115); and 2) many compound verbs have verb—complement relation- ships (208). However, the abundance of homonymous verbs in Persian is problematical, and each verb should be carefully analyzed in accordance with its accompanying arguments. Superficial relationships are often misleading in most examples. Consider the following examples: 36. man dust dar-am 'I have (some) friends' I friend have—I 37. man dust dar-am 'I like (it)' I friend have-I 38. vaqti az dar raft tu hame ehteram-as kard-and when from door went in all respect-him did-they 'when he entered (the room), everyone respected him' 39. vaqti az meydan-e jang dar raft hame xar-as kard-and when from field-E war ran away everyone humiliation- him did-they 'when he turned his back to the battle-field, everyone ostracized him' 40. soma vaqti masjed mirav-i ne—mi-tavan-i zan be-bar-i you when mosque go -you neg—pres-can-you wife/ woman S-take:you 'when you go to a mosque, you cannot take your wife/ a woman with you' 41. game fun baEce-i ne-mi-tavan-i zan be-bar-i you because child-you neg-IND-can-you wife S—take-you 'Since you are (still) a child, you cannot marry' In example (36), dar (from dastan 'to have') is part of a state physical possession predication. In (37), the meaning 104 of the whole predicate is not identical with the accumulated meaning of the constituents. The predication is of a state experiential directed case frame with a covert goal in relation to the experiencer as the filler of the subject slot. In (38), az dar 'through the door' is a path in relation to the predicate raftan go . This is significantly different from (39) in which 93$ is a prefix to raftan 'go' combined to act as a simple verb. In (40), 333 is a so- called 'commitative' or accompaniment case which can be separated from its predicate. Thus, we might have (42) as a paraphrase to (40) but not to (41): 42. soma vaqti masjed mi-rav-i ne-mi-tavan-i zan ba xod be-bar-i you when mosque pres-go-you neg-pres-can- you woman with self S-take-you 'When you go to a mosque, you cannot take a woman/your wife with you' In (41), zan bordan is a compound verb meaning 'to marry' in an action case frame. In the analysis of case frames in the present work, we treat compound verbs as single verbs unless specified otherwise.6 Since our main purpose is to specify predica— tions underlying surface clauses and the nature of predicate- argument relationships, this generalization is not detri- mental though in certain instances further elaborations are to be made. 105 3.1.7.2. Auxiliary Verbs Traditionally, verbs in juxtaposition with nominals in so-called compound verbs were considered to be auxiliaries; the number of such auxiliaries reached as many as twenty- six (Rosen 1898:47-48). However, the distinction between modals, verbs prOper, and auxiliaries was not fully appre— ciated. Marashi (1970) draws distinctions between these categories and finds a number of syntactic and semantic similarities between verbs proper and modals in contrast with auxiliaries. According to him, the modals and auxili— aries are not to be categorized in the same class (422). He assigns the feature [+modal] to tavanest/tavan 'can, may', xast/xah 'would, will, shall', and bayest/bayad 'must' as subclass of verbs (424). On the other hand, dast/dar 'have', bud/ast 'be' and sod/so 'become' are auxiliaries which modify the meaning of the main verb but do not them- selves carry distinct meaning (419). However, like modals, auxiliaries are inflected with respect to number and person. In the following analysis, we usually exclude modals and auxiliaries from the case frame analysis since these are constituents operating at the sentence level and generally they do not affect the case system. However, in examples where their explicit consideration is required we will specify this abstract distinction. 106 3.2. Nominal Features Each noun phrase, when in juxtaposition with a predicate, has certain potentials which differentiate it from other noun phrases. Predicates project features to cases, and cases in turn project features to noun phrases which fill the argument slots. For instance, the predicate tanbih kardan 'to punish' requires that an agent as well as an experiencer be present in the predication. However, the agent by itself projects a [+animate] feature to the noun phrase filling the agent slot. Since the agent should act intentionally, the filler should be capable of acting intelligently. Thus, in a clause such as: 43. madar-am mara tanbih kard 'My mother punished me' mother-my me punishment did-she the feature/+animate/must be attributed to the agent. But, we might have clauses in which the subject slot is not filled by an animate noun phrase. In these cases the agents are metaphorically considered animate, e.g. 44. tabi'at az daijan enteqam mi-gereft nature from Daijan revenge prog-took-it 'Nature was taking its revenge on Daijan' In (44), nature is considered to be potentially capable of performing an action intentionally. 3.2.1. The feature/intent/ Among the six case groups in the present work, only agent (see 3.3.2) is capable of either performing or insti- gating an action. Here, we should distinguish between two 107 types of causation: 1) a causation (with two propositions) in which a causer intentionally causes a performer to act, and 2) a causation (with a single proposition) which is inherent to any transitive predication in an action-process case frame. In other words, an agent performing an action is a causer regarding the performance of that action. Consider the following examples: 45. ahmad xane-ra xarab kard 'Ahmad destroyed the house' Ahmad house-Ach ruined did-he 46. ahmad mar; xandand 'Ahmad made me laugh' Ahmad me laugh (tr) 47. qiyafe-ye masqasem mara be xande v5 dast face—E masqasem me to laughter caused-it 'Masqasem's face/appearance made me laugh' In (45), Ahmad is the agent who acts intentionally and consciously to cause the action. He is simultaneously the performer of the action. In (46), Ahmad is an instigator who causes the occurrence of the action, but is not a performer. Both (45) and (46) can be paraphrased as (48) and (49), respectively. 48. ahmad ba'es-e xarab sodan-e xane sod Ahmad cause-E destruction become-E house became 'Ahmad destroyed the house' 49. ahmad ba'es-e xandidan-e man sod laugh-E 'Ahmad made me laugh' 108 In (47), Masqasem's face/appearance is an instrument which stimulates the occurrence of an action without any intention involved in the predication. Moreover, (45) and (46) can have the following paraphrases, respectively. 50. ahmad amdan xane-ra xarab kard on purpose 'Ahmad destroyed the house on purpose' 51. ahmad amdan mara xandand 'Amad intentionally made me laugh' but we cannot have (52) as a paraphrase to (47): 52. *qiyafe-ye masqasem amdan mara be xande v3 dast Nevertheless, both (45) and (46) can have secondary inter- pretations with no intention involved. Thus, a case frame, which obligatorily possesses the feature intention, requires that the noun phrase filling the subject slot be intelligent. However, in this analysis, the causer is not considered a case in the predicate calculus. A causation associated with a causer is beyond the domain of predication and should be semantically con- sidered as part of prOpositional calculus.7 3.2.2. The featureslinstigator/andeerformer/ Althouth 'causer' is generally considered to be in the propositional calculus (i.e., combination of predications) here, and Egg a part of the case system of the predicate calculus, the distinction between instigation and performance is expedient in the sense that it helps distinguish agent 109 from causer. Both agent and causer have the potential to initiate or incite actions. However, the performer is the one who usually carries out the action and is always the agent where no instigation is present. Nevertheless, in certain reflexive examples, the causer is both the instiga- tor and the performer. Thus, in an example (consisting of two clauses) such as: 53. xod-am-ra vadar be riyazat kard-am self-my-Ach forced to mortification did-I 'I forced myself to undergo mortification' man 'I' is the one who causes himself to perform the action. However, the causer, as one individual, may cause or compel another animate entity to initiate the action. Note the following examples: 54. mard-ha betarafe baq david-and man-pl towards garden ran-they 'The men ran towards the garden' 55. baEEe-ha—ra betarafe baq dav-an-d-and child-pl-Ach towards garden ran-CAUS-past-they 'The children were made to run towards the garden' 56. m3 fahmid-im 'we understood' we understood-we 57. u mozu'-ra be ma fahmand 'He explained the matter to us' he matter-Ach to us understand learn the matter' 58. sobhane-ra xord-am 'I ate breakfast' breakfast- ate-I 110 59. madar—am dava-ra ba zur be man xorand mother-my medicine-Ach with force to me caused to eat 'My mother forced me to take the medicine' (55) and (59) are causative parallels of (54) and (58), (which only have agents), respectively. The derivative causative an on the surface indicates the existence of a causer whether covert on the surface as in (55) or overt as in (59). However, an is not only a causative marker, but also an indicator of transitivity with certain simple verbs whose base forms do not allow patients8 (Khanlari 1973: 258-59). Example (57) is a corresponding transitive clause to (56) since 23 in (57) is not the performer of the action. Causation (in the propositional calculus) might also be represented by prepositional phrases which show the peripher— ality of causer to the predication more clearly. Consider the following example: 60. be dastur —e aqajan masqasem nardeban-i avard to order-E Aqajan Masqasem ladder-a brought-he 'lit. with the order of Aqajan, Masqasem brought a ladder' This sentence could have the following semantic representation: 61. [aqéjan be masqasem dastur dad] masqasem nardeban— i avard] Aqajan to Masqasem order gave Masqasem ladder-a brought We might also select a causer as the subject of the causative sentence (with two underlying predications) if 111 the verb vadar kardan 'force/oblige' constitutes the verb of the matrix sentence and if an embedded clause dominated by it includes an agentive case in its underlying representa— tion. 3.2.3. The feature/potent/ By potent we refer to the process of giving prominence to a case beyond its domain so that its valence extends to the realm of another case. This process (elaboration) devaluates the potency of one case and assigns greater power to another case which usually stands as a representation of a different case. In other words, potency refers to the semantic specification of a case if that case happens to be capable of gaining potency. Following Chafe (1970:109)’ we differentiate, for instance, between a potent agent and a real agent. Note the following examples. 62. bad 0 baran xane-ye ma-ra xarab kard-and wind and rain house-E we—Ach ruined did-they 'The rain and wind(together) demolished our house' 63. javab-e madar—am aram-am kard answer-E mother-my calm-me did 'My mother's answer cooled me off/calmed me' 64. bargastan-e masqasem reste-ye afkar-am-ra pare kard return—E Masqasem ties-E imagination-my-Ach torn did 'Masqasem's return disturbed the train of my thought' 65. yek vaqe'e-ye nagahani bar 35 xosk—am kard one event-E sudden on place dry-me did-it 'An unexpected event shocked me' 112 66. Eesm-ha-ye leyli mar; ta'qib mi-kard eye-pl-E Leyli me follow prog-did 'Leyli's eyes were chasing me' Here, in (62) bad 0 baran is not a real agent, but rather an instrument which has been promoted to the subject position by assuming that it is inherently potent to cause damage and destruction. javabe madaram 'my mother's answer' in (63) and Masqasem's return in (64) also possess this feature. Leyli's eyes in (66) is an instrumental case being promoted to a subject position, suppressing the agent and gaining potency. This clause can be expanded into (67): 67. leyli b5 cesm-hE-ye xod mara taqib mi—kard Leyli with eye-pl-E self me follow prog-did 'Leyli's eyes chased me' 3.3. Nuclear Cases Nuclear cases or roles are necessary constituents of predications. In other words, their occurrences are necessary to differentiate one case frame as opposed to others. On the other hand, peripheral roles are not diag— nostic and their presences are not required as necessary components of predications. Among the possible cases within a predication in Persian, some are nuclear and crucial in meaning interpre- tations, whereas others are peripheral and non—essential. Here, we should emphasize the differentiation between surface and deep realizations. The number of arguments 113 at the surface need not correspond to the number of cases in the semantic representation on a one-to-one basis. Nuclear roles (Cook's 'propositional roles') are part of the valence of the verb, whereas peripheral roles (Cook's 'modals') are optional entities which might expand the meaning but are not relevant to the specification of underlying predi- cation. Longacre, inspired by formal logic, makes a distinction between a propositional calculus—- a combination of predica— tions-- and predicate calculus, which involves predicates and their basic accompanying arguments (1976:98). Pike also makes a similar distinction in his argument concerning nuclear versus marginal tagmemes. To illustrate it, Pike differentiates sc0pe (nuclear) and location (marginal) in relation to the predicates, the latter of which is weakly related to the predicate and only offers free information (1977:47-48). Pike's sc0pe covers goal and locative in Longacre's terminology, and Pike's location roughly corresponds with 'place' as it will be differentiated from 'locative' in the discussion on section 3.3.5 of the present work. In this section,we categorize nuclear roles into six categories, each of which has certain subcategories. 3.3.1. 'Direction' Group In this group, we include cases to which the action is directed, whether or not the receivers perceive the action 114 mentally, undergo the action physically, or are simply goals. 3.3.1.1. Experiencer. By experiencer, we refer to an animate entity who phy- sically or mentally perceives the effects of an action or an environmental condition. Here, we differentiate between patient and experiencer in the sense that the experiencer must be animate and should not undergo physical change or spatial dislocation. The expiencer's distribution is res- tricted to experiential case frames (see 3.5.1-3.5.6 below). The experiencer should be either an animate entity or an inalienable possession related to an animate entity. The latter is a potent substitute for the former. Consider the following examples: 68. hayajan-e asdolamirza be hame asar kard excitement-E Asdolamirza to all effect did 'Asdolamirza's excitement influenced everyone' 69. daijan be u tasar mi-zad Daijan to he curse prog—hit 'Daijan used to shout threats at him' 70. aqa goft-and be—rav-i anja Aqa said-they/he S-go-you there 'Aqa said (to me) that you should go there' 71. ba cesm-ha-ye leyli modati harf zad-am with eye-pl-E Leyli some time speech hit-I 'lit. I talked to Leyli's eyes for some time' be hame in (68), and be u in (69) are experiencers who are overt on the surface. The experiencer in (70) is partially 115 covert. However, one can trace an experiencer in the under- lying semantic configuration, expressions such as be man/u/ mi 'to me/him/us' and so on. In (71), the real experiencer is Leyli, however, Eesm-ha'eyes' as inalienable possessions seems to be the target. The most unmarked marker of experiencer is be; followed by a nominal. ;£é -- the accusative marker-— and experiencer markers are mutually exclusive. According to the available data, experiencers are marked in action—process case frames. The other markers are a; 'from', b: 'with', and baraye 'for'. 35 is restricted to roles which are both experiencer and source. b: and baraye only occur with sensation and speech predictions (see 3.5.6). Consider the following examples (roles in parentheses correspond with the underlined grammatical forms henceforth). 72. so'alat-e saxti az man mi-kon—ad (experiencer) questions-E hard from me pres-does-he 'He asks me/is asking me difficult questions' 73. yek so'al-i mi-xah—am az some be-kon-am (experiencer) one question-a pres-want-I from you S-do-I 'I want to ask you a question' 74. dastan-e xarab‘sodan-e otomobil-ra baraye ma hekayat kard (experiencer) story-E ruined become-E car-Ach for us story did-he 'He told us the story of how the car had broken down' 75. az post-e dar ba asdolamirza sohbat kard-am (experiencer) from behind—E door with Asdolamirza speech did-I 'I talked to Asdolamirza from behind the door' 116 76. b; faryad mara seda zad (exper) with shout me noise hit 'He called me as he shouted' 77. dast-i be sane-ye man zad (exper) 78. az post-e dar asdolamirza-ra moxatab qarar dad-am (exper) from behind-E door Asdolamirza-Ach addressed placed gave-I 'I addressed Asdolamirza from behind the door' 79. *az post-e dar b5 asdolamirza—ra moxatab qarar dad-am Experiencer is dominated by an instigator in the action- process case frames. It might fill the subject slot in state or process experiential case frames. Moreover, a body part (such as mind, heart, etc.) may substitute for it as an allo-experiencer with the suppression of the real experi— encer. Consider the following examples: 80. baraye pedar-am na-rahat-am fbr father-my not-comfortable-I am 'I am anxious for my father' 81. xater-e man saxt masqul bud mind-E I hard busy was-it 'I was very busy' 82. man qam-e xod—r3 faramus kard-am I grief-E self-Ach forget did-I 'I forgot my own grief' 83. man az in damame mi-tars-am I from this demon pres—fear-I 'I am scared of this demon' 117 84. asdolamirza dast asabani mi-sod Asdolamirza prog angry prog-became 'Asdolamirza was getting angry' 85. modati tulani fekr kard-am time long thought did-I 'I thought for a long time' 86. del-e masqasem be rahm amad heart-E Masqasem to pity came—it 'Masqasem felt sorry' In example (80), the grammatical subject man 'I' (experiencer) has been optionally omitted according to a general Persian rule which gives the speaker the choice not to choose the subject pronoun. (81) is different from (80) in that even if 233 'I' as the real experiencer is optionally chosen, it has lost its semantic valence. The verb does not agree with the subject. In (82), man is both agent and experiencer. The underlined nominals in (83) and (84) are experiencers. The experiencer may surface as subject of intransitive and transitive clauses; as indirect object (in action-exper- iential completable case frames), and direct object (action— process experiential directed case frames) of di—transitive clause roots; and as subject of descriptive clauses (state ambient experiential case frames) (see 3.8 for details). An agent in an embedded clause can be represented as an indirect object in a main sentence with the function of experiencer. For instance, example (87) 87. [[aqa goft—and] [man be soma mi-gu-yam]] Aqa said-he I to you pres-say-I has (88) as its parallel paraphrase: 118 88. aqa be man goft—and be soma be-gu—yam Aqa to I said-he to you S-say-I 'Aqa told me to tell you ...' 3.3.1.2. Patient Patient refers to an animate or inanimate entity who 1) physically undergoes change of condition, 2) physically relocates its position under an external pressure, and 3) stands in a possession relationship with another entity with- in the same predication. The patient is not restricted to [+ affected] entities but it might indicate entities possessed. It can be used in collocation with either physical, directional, and/or possession state, process, action-process predications, though rarely with action predicates. Consider the following examples. The underlined grammatical forms are semantically deep patients. 89. dar—ra post-e sar-e va'ez bast-am door-Ach behind-E head-E preacher closed-I 'I closed the door when the preacher left' .90. sar-es tuye puze-ye pelang gir karde bud head-his in mouth-E leOpard caught done was-it 'His head was stuck in the leopard's mouth' 91. xod-am tuye. qabr mi-gozar-am-EE self-my in grave pres-put-I-you 'I'll bury you' 92. belaxare man be zamin oftad-am finally I to ground fell-I 'Finally, I fell down' 119 93. haftir-as az dast-as oftad revolver-his from hand-his fell 'His revolver fell from his hand' 94. mar; bord ta sangar-e xod-eman me took-he upto trench ourselves 'He carried/took me to my own trench' 95. man esq-am-ra az dast mi-dad-am I love-my-Ach from hand prog-gave—I 'My love was being taken from me' 96. yek teke kohnefiye siyah sode az zamin qapid-am one piece cloth-E black become from ground grabbed-I 'I snatched a piece of black cloth from the ground' In example (89), man 'I' as an agent to the predication has been optionally omitted from the surface structure. In example (90), a pronoun 3 'he' might be used but it would be semantically void since its presence or absence does not change the basic meaning. In order words, the patient has been brought into focus and the original subject is supressed. In the rest of the examples above, the patient is the entity which is grabbed (96), lost (95), or relocated (92,93,94). -£§ is the only postposition which occurs in juxta- position with patient if a postposition is required. As the data shows, the patient takes the postposition -£§ in action—process case frames, whereas it is unmarked in state, and process predications. The reason why the patient is unmarked in state and process case frames might be that patients are not affected by the performer's activity in the same manner as they are in action-process case frames. 120 However, the exact status of :gé vis a via the patient is controversial for the following reasons: a. :£§_can occur with range and experiencer as well, and b. :35, as Windfuhr states, "... marks the direct object if the latter [i.e., the direct object] is 'definite' either implicitly by context or explicitly by the presence of any kind of pronoun or a name..." (1979:50). Furthermore, Windfuhr claims that there is sufficient evidence that the indirect object can also be marked by the postposition :5; (1979:50). The data show that there are some exceptions to what has already been mentioned. That is, there are examples in the action-process column which are expected to be marked with the postposition :3é, but which are actually unmarked. Note the following examples: 97. m3 b3 tofang mi-zan-im miyan-e qalb-es we with gun pres-hit-we canter-E heart-his 'I'll shoot him right in the heart' 98. *m3 b5 tofang mi-zan-im miyan-e qalb-es;£§ 99. m3 b5 tofang miyan-e qalb-e§;£é mi-zan-im 100. (?)m3 b3 tofang miyan-e qalb-es mi-zan-im 101. masqasem cand gilas ruye miz gozast Masqasem a few glass on table put-he 'Masqasem put a few glasses on the table' 102. (?) masqasem Eand gilas-ra ruye miz gozast 103. masqasem gilas-ha-ra ruye miz gozast 121 104. * masqasem gilas-ha ruye miz gozast 105. xod-am tu-ye qabr mi-gozar-am-et self-my in grave pres-put-I-you 'I'll bury you' 106. *xod-am tuye qabr mi-gozar-am-at-ra 107. xod-am to-ra tuye qabr mi-gozar-am you-Ach 108. *xod-am to tuye qabr mi-gozar-am Here, it seems, word order and clitic formation are signi- ficant with respect to the occurrence of :ré. In (97), the word order is 3V0, thus no :3; occurs. On the contrary, in (99), the word order is SOV (i.e., subject+object+verb the most unmarked), thus IE; occurs. (98) and (100) are odd or unacceptable probably because one does not expect -ra in (98), but -ra is required in (100). The partitive £323 'a few' in (101) versus its non—occurrence in (103) eliminates Zgé selection. The word order in (105), as in (97) is SVO. Thus IE; does not occur. However, (107) as a paraphrase to (105) is marked by the postposition iii: The patient is dominated by an instigator in action- process case frames. The noun phrase acting as the patient in the state and process columns fills the grammatical subject slot in physical and physical locative case frames. Thus the occurrence of patient in action-process predications implies an agentive case in the case frame. The patient is used as a subject in intransitive clauses representing state and process physical and physical measurable 122 case predications. The patient surfaces as subject in descriptive and equative clauses (see 3.8 for details). 3.3.1.3. 2231 Following Longacre (see 2.1.2.9 of the present work), we assign goal to cases towards whom an action is directed, for whom an action is performed (cf. c.c.l.4 below), and to an animate entity who non-transitionally possesses an entity. Here, we distinguish between a patient, which is affected by the predication and undergoes a physical change of state, and a goal as a locale towards which a predication is dir- ected with no impact on it whatsoever. Consider the follow- ing examp188where the underlined entities are underlying goals in their related predications: 109. az ruye taxt be zamin oftad from on bed to ground fell-he 'He fell off the bed' 110. vasat-e an barun—e golule xod—es-ra resand be ma middle-E that rain-E bullet self-him-Ach reached to we 'As the rain of bullets continued, he reached me' 111. masqasem mara betarafe yek semsad-e bozorg hol dad Masqasem me towards one box-tree-E big push gave—he 'Masqasem pushed me towards a big box—tree' 112. dustali sise-ye dava-ra be qasd-e sar-e sazde Dustali bottle-E medicine-Ach to destination-e head-E partab kard prince threw 'Dustali threw the bottle of medicine at the Shazde's head' 123 113. dozd-ha rixt-and tuye xeyme-ye ma thief-p1 poured-they in tent-E we 'The thieves attacked our tents' 114. anha sarnevest—e qamangizi dast—and they destiny-E sad had-they 'They all faced the same sad fate/end' 115. man esq-am-ra az dast dad-am I love-my-Ach from hand gave-I 'I lost my love' 116. man yek bastani baraye leyli xarid-am I one ice—cream for Leyli bought-I 'I bought Leyli an ice-cream' In (115), man 'I' is the entity from whom something has been taken away. Sullivan (1980) uses a maleficiary case as the specification of an entity who suffers loss from the act (309). This term is much more elaborative and suggestive than 'goal' in particular cases such as that in (115) uses here. However, in order to be more consistent with the original theoretical framework, we still follow Longacre's more general terminology. The most unmarked goal-marker in Persian is 23:, which occurs in practically all case frames where goal exists as a case; it does not occur where certain additional specifications are implied or required. Other markers, in a relative hierarchy of distributiong, are betarafe 'towards', baraye 'for', £213 (lit. in) 'towards', and nesbatbe 'in I relation to. Consider the following examples: 124 117. masqasem raft tuye sarh-e jang-e kazerun Masqasem went in description-E war-E Kazerun 'Masqasem started explaining about the Kazerun battle' 118. ketab-ra partab kard tu/be/betarafe sar-am book-Ach threw towards head-my 'He threw the book towards me' 119. leyli betarafe/*be dar-e xane-ye xodesan david Leyli towards door-E house-E themselves ran 'Leyli ran towards the door of their house' 120. m5 b3 lebas-e xab betarafe/* be u david-im we with dress-E sleep towards he ran-we 'We ran after him in our night gowns' 121. man baraye doxtar-ha sokolat xarid-am I for girl-p1 chocolate bought-I 'I bought the girls (some) chocolates' 122. mi-xah-am az some yek Eizi be-pors-am pres-want-I from you one thing S-ask-I 'I want to ask you a question' In example (118), £3, be, and betarafe can be used, but the related predications are not paraphrases. Tu, and be imply that the target was hit, whereas betarafe emphasizes more the direction of the predication rather than the target itself. In (119) and (120), be and betarafe are not inter- changeable. In (121), doxtarha 'girls' is a beneficiary. In (122), the nominal following 2 has double functions; 'you' is goal, experiencer, and source as well. V . soma Goal occurs as margin as scope in intransitive clauses (action physical motion case frames). It underlies the 125 subject slot in certain transitive clauses (representing process physical possession directed case frames). Goal coreferential with agent surfaces as subject in action- process directed case frames. In ditransitive clauses, goal underlies indirect objects (see 3.8 for details). 3.3.1.4. Client Longacre does not distinguish a client from goal. However, in the process of analyzing the Persian material within a case grammar model, some additions seemed necessary. Sullivan (1980) uses client as an animate entity "for whom the act is performed" (309). In this analysis, we use the term 'client' to refer to certain special cases. Sullivan's terminology includes Longacre's goal, and Cook's and Chafe's beneficiary. However, none of Sullivan's examples offer instances in English where a beneficiary is differentiated from a client (310). Here we classify client as a case in a predication where an action is performed or a condition is stated for some animate entity without him/her being an experiencer to feel the action, or without being the locale towards whom an action is directed, or without being an animate to profit from the action. We particularly dis- tinguish between the following cases. 123. del-am baraye u sur mi-zan—ad heart—my for he anxiety pres-hit-it 'I feel anxious for him' 126 124. baraye pedar-am na-rahat-am for father-my not-comfortable—I am 'I am anxious for my father 125. man mandana-ra dust dar-am I Mandana-Ach friend have-I 'I love Mandana' 126. man baraye hasan k3: mi-kon-am I for Hassan work pres-do-I 'I work for Hassan' 127. name-i baraye ali nevest-am10 letter-a for Ali wrote-I 'I wrote a letter to/for Ali' 128. sa'at-ha baraye esq—am gerye kard-am1 hour-p1 for love-my tear did-I 'I wept for my love for hours' 129. ketab—ha-ra az dast-e ali qapid-am book-pl-Ach from hand-E Ali grabbed-I 'I snatched the book from Ali's hands' 130. ketab-ra baraye ali post kard-am book-Ach for Ali mail did-I 'I mailed the book to/for Ali' In (123) and (124), the underlined nominals are clients since 1) no action is directed towards them, and 2) they do not benefit from the statement or the process involved. In (125), man 'I' is an experiencer, whereas Mandana is a goal. In (126), (127), and (128), the nominals are beneficiaries. In (129), Ali is the source. In (130), Ali is the goal/ beneficiary. Beneficiary, goal, and maleficiary can be 127 conflated into 'goal' since they are affected by the action in one way or another. However, 'client' is differentiated from them by the fact that it is not in physical or sensical relation with the patient or experiencer. The unmarked client-marker is baraye 'for', which represents it on the surface. 3.3.2. Agent The agentive case identifies the performer of an action. It marks the animate entity who instigates and performs the action, or the inanimate entity which has potential to cause processes or deformation, or the animate entity which exerts energy, even though its effect does not pass to other entities. Causer, which is not considered a case in predi- cate calculus because it occurs in constructions consisting of two propositions, is different from the agent in the sense that agent is attributed with the features [+instigator] and [+performer] whereas the causer is [-performer] and [+instigator]. Consider the following examples where examples 131, 133, and 135 are single clauses; and 132, 134, and 136 are two clauses each. 131. hasan xane-ra xarab kard Hassan house-Ach ruined did 'Hassan demolished the house' 132. ali hasan-r3 vadast xane-ra xarab kon-ad Ali Hassan-Ach forced house-Ach ruined do—he 'Ali made Hassan demolish the house' 128 133. hasan xabid ruye kanape Hassan lay on sofa 'Hassan lay down on the sofa' 134. hasan be baradar-as bazur qaza xorand Hassan to brother-his with force food made eat 'Hassan made his brother eat food' 135. hasan david Hassan ran 'Hassan ran' 136. mo'alem-e varzes hasan-r3 davanid teacher-E sport Hassan-Ach made run In example (131), Hassan is an agent who initiates the action of demolishing the house. In (132), (134), and (136), All Hassan, and teacher are causers. (134) and (136) can have (137) and (138) as their corresponding paraphrases, respec- tively. 137. hasan vadast baradar-as qaza be-xor—ad Hassan made brother-his food S—eat—he or hasan baradar-as-ra vadast qaza be-xor-ad Hassan brother-his-Ach forced food S-eat-he 138. mo'alem-e varze§ vadast hasan be-dav—ad teacher-E sport forced Hassan S—run-he Here the verbs are superficially in their grammatical sub- junctive forms. Thus, if causer occurs at the propositional calculus level, the original performer is the agent of the subordinate predication. 129 Intention as a feature is very significant here, since an instigator without intention is a stimulus acting as an instrument. Note the following examples: 139. nane-as mars mi-tarsan-ad mother-his me pres-scare-she 'His mother scares me' V V - V '- 140. nane-as hamise mara az post-e sar mi-tarsan-ad mother-his always me from behind-E head pres— scare-she 'His mother always scares me from behind' 141. taze vared mara zahre tarak kard new arrived me gall-bladder break did 'The new comer frightened/scared me' 142. in abavi-ye gem; ham har ja mi-ras-ad mara zahre tarak mi-ko n-ad this father-E you too any place pres-reach-he me scares 'Your father scares me wherever he sees me' In (139) and (141), the underlined nominals are instruments since their presence is the means by which someone becomes scared or terrified. His mother in (140) and your father in (142) are agents who intentionally initiate the actions. Agent might be coreferential with source, goal, or experiencer, and in special examples, coreferential with the patient. Consider the following examples. Roles related to the underlined nominals appear in parentheses. 143. man xodzra az panjere—ye kuéak be dorun—e sanduqxane andaxt-am I self-Ach from window-E small to inside—E warehouse threw-I (A/P) 'I jumped into the warehouse through the/a small window' 130 144. man xod-r3 be an taraf kesand-am (A/P) I self-Ach to that side pulled-I 'I went there' 145. daijan tofang-r3 part kard (A/S) Daijan gun-Ach threw 'Daijan threw the gun' 146. zarbgir avaz mi-xand (A/S) drum player song prog-sang 'The drum player was singing' 147. masqasem Eesm be qeyEi-ye baqbani duxte bud (A/Exper) Masqasem eye to scissors gardening fixed was-he 'Masqasem was looking at the garden shears' 148. man yek qamus-e bozorg xarid-am (A/G) I one dictionary big bought-I 'I bought a dictionary' Agent dominates instruments if a patient is present in the predication, whether implicitly or explicitly. This implies that if there is an instrument in an action-process case frame, and a patient too, the occurrence of an agent is a necessity. The agent is the preferred filler of the subject position in action and action-process predications which underlie transitive, intransitive and ditransitive clauses. As mentioned above, the agentive case never takes a marker in a simple clause. In causation, where a new instigator is introduced, the agent in the embedded clause takes an accusative form on the surface. The so-called 'passive' clauses are not successful in Persian if agent is specified. However, if the agent is required, it is marked by bevasile 'by means of', be dast 131 'by', etc. with minor differences in meaning. Thus, the hierarchy of acceptibility for an equivalent predication of the English clause The innocent linguist was killed by the guerrillas would be as follows: 149. a. Eerik-ha zabansenas-e bi-gonah-ra kost-and guerrill-pl linguist-E not-guilt-Ach killed- they b. zabansenase bigonahra kostand (unknown agent) zabansenase bigonah koste sod d. zabansenase bigonah bevasile-ye cerikha koste sod by 3.3.3. 'Giver' Group The entities in this group are either the locales from which an action originates (source) or a locale through which an action passes or an entity is transferred (path). 3.3.3.1. Source Source refers to 1) an entity from which a physical sensation emanates. The following underlined nominals are sources: 150. negah-e cesm-ha-yg siyah-as ba negah-e man talaqi kard look-E eyes-pl-E black-her with look-E I collision did 'We looked at each other in the eye' 151. seda-ye Eakos-e dar-e xane boland sod sound—E knocker-E door-E house loud became-it 'The knocking on the door became louder' 152. 153. 132 bu-ye 21152 be masam rasid odor-E onion to sense reached 'One could smell the onion' sedE-ye sirali qasab senide sod voice-E Sirali butcher heard became 'Sirali the butcher was heard' 2) the locale from which a predication starts with the intention of moving to another direction or terminating in another locale, e.g. and 154. 155. 156. 157. az taxt oftad-am from bed fell-I 'I fell off the bed' mesle baran golule az in taraf va an taraf mi-barid like raining bullet from this side and that side prog-rained 'Bullets were raining everywhere' az sat-es ta dam-e naf-as-ra borid-am from head-his to edge-E navel-his-Ach cut—I 'I cut him in half from his head to his navel' dar paseband az in dande be an dande mi-sod-am in gazebo from this rib to that rib prog-became-I 'I was tossing and turning under the gazebo' 3) the non-transitory owner who gives away his property, e.g. 158. 159. man meqdari az bastani-ye xod-r3 b3 meyl be leyli dad-am I some from ice-cream-E self-my with eagerness to Leyli gave-I 'I gave some of my ice—cream to Leyli willingly' dustali esq-am-ra az man gereft Dustali love—my-Ach from I took 'Dustali stole my love from me' 133 160. masqasem az dast-e 3 farar kard Masqasem from hand-E he escaped did 'Masqasem ran away from him' In examples (150) to (153), either a sound wave, a light wave, or an odor originates from an entity (i.e., source) which is experienced by an animate entity (experiencer). Examples (154) to (157) exemplify the occurrences of locales from which the predications originate and are directed to other arguments. Examples (158), (159), and (160) indicate the original owners before the possession is taken away. In. Sullivan's (1980) terminology, the last group of entities are maleficiaries. Since source is pertinent only when some kind of action takes place, it is restricted to process, action-process, and action predicates. In other words, an odor is never smelled unless it is given off by a source and experienced by an experiencer. In (158), source has no semantic valence unless the action starts from a locale and moves towards another locale. Source is marked by 33 except in sensation case markers (see 3.5.6), where the source and its range are in part- whole relationships. In this case no marker appears on the surface. However, in related paraphrases, the marker can optionally appear. Consider the following examples: 150a. negah-i ke az cesm-ha-ye siyah-as bar mi-xast b5 _ ffgm arose negah—e man talaqi kard 134 1513. seda-i gg Eakos-e dar-e xane boland sod The source may be coreferential with the agent. Consider some examples: 161. zarbgir avaz mi-xand (A/S) 'The drum player was singing' 162. 595 goft—and be-rav-i anja (A/S) 'Aqa said that you should go there' A source in collocation with range may act as a potent agent, that is, an entity which has the potential to cause processes or even to affect other arguments in action-process predications. The following examples clarify this point: 163. lavab—e madar-am aram-am kard (R-S) answer-E mother-my calm-me did 'My mother's answer cooled me off' 164. hayajan-e asdolamirza be hame asar kard (R-S) excitement-E Asdolamirza to all effect did 'Asdolamirza's excitement influenced everyone' Source may represent underlyingly a surface margin in intransitive clauses (action physical motion case frames). It may also surface as a subject in transitive and ditrans- itive clauses when it is coreferent with an agent (for more details see 3.8). 3.3.3.2. Path Path is the locale through which motions, and actions are performed. In other words, the motion transverses through a path in a particular direction. Note the following examples: 135 165. daijan az post-e eynak-e dudi-ye hamisegi-yas negah mi-kard Daijan from behind-E glasses-E smoky—E permanent- his was looking 'Daijan was peering through the thick dark glasses that were his trademark' 166. hata fekr-e yek zan-e digar az maqz-am na-gozast even thought—E one woman-E other from mind-my not-passed 'I never thought of any other woman' 167. arus o damad az tariq3e neysabur be mashad raft-and bride and bridegroom from way-E Neysabur to Mashad went-they 'The newly married couple went to Mashad via Neyshabur' The path is marked in physical motion clauses. It is marked by 35 or by prepositional phrases such as az tariqe 'via', az vasate 'through', az rahe 'through, via', az kenare 'along', tavasote 'by', etc. However, where path is core- ferential with the agent, it is unmarked. The path also refers to the temporary owner of an entity in transfer and acquisition verbs. However, in ex- amples where a path, a goal, and an agent occur with transfer, grab, and acquisition verbs, Persian shows a proclivity to use co-ordinated clauses rather than simple clauses. In other words, a collocation of these entities within a simple predication normally surfaces in two clauses with a con- junction or with an embedded clause marked by kg. Consider the following examples: 136 168. man lebas-ha-ye xoskeli baraye doxtar-am xarid-am I dress-pl-E beautiful for daughter-my bought-I 'I bought beautiful dresses for my daughter' 169. yeki az qgmoxis-ha hand sise odokolon baraye daijan xaride bud one from relative-pl a few bottle eau de cologne Daijan bought was-he 'One of the relatives had bought a bottle of eau de cologne for Daijan' 170. (?) man qalam-ra dad-am be hasan baraye to I pen-Ach gave-I to Hassan for you 'I gave Hassan the book for you' 171. qalam-ra dad—am be hasan ke be-de be to pen-Ach gave-I to Hassan that S-give to you 'I gave Hassan the book for you' 172. (?) masin-ra be ali baraye zan-as foruxt-am car—Ach to Ali for wife-his sold-I 'I sold the car to Ali for his wife' 173. (?) ali masin-ra baraye zan-as az man xarid Ali car—Ach for wife-his from I bought 'Ali bought the car from me for his wife' 174. ali misin-e mara baraye zan—as xarid Ali car-e my-Ach for wife-his bought 'Ali bought my car for his wife' In (168) and (169), I and one of the relatives are transitory owners. The grammatical forms represent paths coreferential with agents. (171) includes a matrix clause with an embedded clause related to it by the particle kg 'that'. (172) involves an agent, a path, and a goal. The cases in (173) are an agent/path, a goal, a patient, and a source. Source in (173) is an agent which has been demoted to source position 137 and marked by 35. Both (172) and (173) seem odd, though acceptable.12 (173) is relatively more acceptable than others since All is an agent/path and his wife is a goal who will be the non-transitory owner of the property. The occurrence of path is common with verbs representing grab, transfer, and acquisition predications, as in case frames in groups G' and H' (see 3.5.10; 3.5.12). 3.3.4. Instrument Instrument refers to an entity which is used by an animate actor to carry out some action or instigate a pro- cess. Instrument is inherent in predications in which an agent intentionally performs an action to affect a patient or to instigate a process no matter whether the instrument is covert or overt in the surface structure. It is apparent that any agent who instigates an act which is to affect someone or something uses an instrument. However, the occurrence of instrument is not restricted to action-process predications, but may optionally or even obligatorily occur in state and process case frames. Note the following examples where the underlined nominals are instruments. 175. ha angost -an—e lagar-e xod dar-e xane-ra nesan dad with finger-pl-E thin-E self door-E house-Accm show gave 'He pointed to the door of the house with his skinny fingers' 138 176. ba bil zad tuye sar-am with shovel hit in head-my 'He hit me on the head with the shovef 177. b5 yek golule zad tuye sar-es with one bullet hit in head-his 'He shot him in head with a bullet' 178. baEce-ha-ye seytun dar-e otaq-ra surax kard-and child-pl-E naughty door-E room-Ach hole did-they 'The naughty children made holes in the door' 179. dat-as-ra ba Eagu borid hand-his-Ach with knife cut-he 'He cut his hand with a knife' In (178), the instrument does not surface but is implied in the context. It is obligatorily present in the underlying representation. The instrument may be an emotional stimulus which brings about certain conditions or triggers an action. In examples of this type instrument is not marked by bi. Consider the following examples. 180. u az tars dandan-ha-yas gofl sode bud he from fear tooth—pl-his locked became was 'He was so scared that he couldn't talk' 181. seda-yas az xasm mi-larzid voice-his from anger prog—trembled 'He was trembling with anger' 182. daijan az adam-e movafaqiyat-e xod saxt gerefte bud Daijan from lack-E success-E self hard depressed was 'Daijan was depressed at his lack of success' 139 Note that az tars 'from fear', az xasm 'from anger', and so on are stimuli which cause someone or something to function or malfunction. Moreover, an animate entity who unintentionally triggers a condition or causes the initiation of a process is an instrument rather than an agent (see also 3.3.2 in this chapter) e.g.: 183. qahreman-e jang-e kazerun dast az tars-e dozd sekte mi-kard hero-E war-E Kazerun past prog from fear-E thief shocked prog-did 'The hero was shocked by the sight of the burglar' 184. hame-ye mard-ha-ye irani az madar zan-esan mi-tars—and all-E man—pl-E Iranian from mother wife-their pres-fear—they 'All Iranian men/husbands are scared of their mother-in-laws' The most unmarked instrumental marker is £3 'with'. B; and 3513 alternate in source-range relationships as in (185), (186), and (187) below. However, for inalienable possessions, az as a marker choice is eliminated. 185. ba/az faryad-e madar-am az xab bidar sod—am with/from shout-E mother-my from sleep awake became-I 'I woke up with my mother's scream' 186. sobh-e zud az/ba seda-ye zang az xab parid-am morning-E early from/with voice-E ring from sleep jumped-I 'I woke up with the sound of the bell' 140 187. balaz xornas-e u tamam-e ahl-e sahr bidar sod-and with/from snoring-E he all-E inhabitant-E city awake became-they 'The people of the city were awakened by his snoring' 188. b3/*az angost-e xod mara nesan dad with/from finger-E self he point gave 'He pointed at me with his finger' The emotional stimulus is exclusively marked with az. e.g.: 189. az avval az esg mi-tarsid-am from beginning from love IND-fear-I 'I was scared of love from the beginning' E2 as a variant of 25 is less common. However, a few examples such as: _ [ha] _ _ _ 190. yek ruz be Eesm-e xod-eman ezdaha-ra did-im one day with eye-E self-our dragon-Ach saw-we 'One day, I saw the dragon with my own eyes' Instrumental case underlies margins cooccurring with agents in transitive clauses. It is obligatorily present in transitive clauses representing underlying action-process physical instrumental case frames. 3.3.5. Locative Here, we posit 'place' as a peripheral case and locative as a nucleus case. Locative is obligatory to the predication, whereas place is beyond the domain of predicate calculus. Place can optionally occur with most case frames in all parameters, whereas the occurrence of 141 locative is restricted to locative case frames (see 3.5.9). Consider the following examples: 191. dustali dar alam-e ro'b o vahsat bud (locative) Dustali in world-E fear and fear was-he 'Dustali was in a world of fear' 192. xod-am goldan-ra tuye qafase gozast-am (locative) selfemy vase-Ach in shelf put -I 'I myself put the vase on the shelf' 193. aqajan mehmani-ye bozorg-i dar xane-as dad (place) Aqajan party-E big-a in house-his gave 'Aqajan gave a party in his house' The occurrence of locative nominals in (191) and (192) are obligatory whereas in (193), it is optional. The locative is commonly marked with $33 'in'. Tu is a variant of iii: Other markers are 22 as a variant of bar on and prepositional phrases such as dar beyn 'between', dar moqabel 'in front of', etc. The locative should be distinguished from path, source, and goal, which most often replace it. The locative indicates the place where the predication occurs, and is different from source, path, and goal due to the fact that no impli- cation as to the direction of the action to or from the locale can be obtained. The locative case surfaces as a nuclear tagmeme in transitive and intransitive clauses representing action locative and state locative case frames, respectively. 142 3.3.6. Neutral Group In this category we include range, identifier, and measure. All of them, to a certain extent, complete the predication, specify the predication, or stand in an equative relation with other notional categories. 3.3.6.1. Range Range refers to 1) the specification of a predicate. It refers to any nominal on the surface which completes the meaning of the predication without the nominal itself being the target of the action. Consider the following examples: 194. man qiyafe—ye daijan-ra ne-mi—did-am I mien-E Daijan-Ach neg-IND-see-I 'I couldn't/was not able to see Daijan's face' 195. sobhane-ra xord-am breakfast-Ach ate-I 'I ate breakfast' 196. seda-yeparande-ha-ra mi-sanid-am voice-E bird—pl-Ach IND-heard—I 'I heard/could hear the birds' singing' 2) the by-product of an activity of a predication in physical action frames. Thus, in examples such as: 197. pedar-e daijan haft emarat saxte bud father-E Daijan seven building built was 'Daijan's father had built seven buildings' 198. pas az do 321 leyli do baEEe zaide bud after from two year Leyli two child borne was-she 'After two years, Leyli had given birth to two children' 143 the underlined nominals are ranges. Here, we should emphasize the fact that in range-source juxtapositions, range must occur in the surface structure of Persian. Thus, in 199. bu-ye piaz mi-sanav-am odor-E onion pres-hear—I 'I smell onion' the range cannot be omitted. Range commonly occurs with predicates implying factual knowledge, with sensation and speech predications (see 3.5.6), and with physical action predicates. Range should be differentiated from patient and goal as they were discussed in 3.3.1.2 and 3.3.1.3. Range surfaces as object as complement in transitive and ditransitive clause roots (see 3.8 for further details and examples). 3.3.6.2. Measure Measure is a case representing underlyingly the surface nominals which quantify a predication. Quantification may also include prices. Measure is a nucleus in the predica— tions in which the predicates are attributed with the feature measurable. Take the following examples into consideration: 200. xeyli baEEe 331 bud-am very child year was-I 'I was very young' 144 201. fasele-ye ma ta dosman sad qadam bud distance-E our to enemy 100 step was 'The distance between us and the enemy was 100 steps' 202. az vaqti baEEe-ha-yam rafte -and bist pond kam karde-am from time child—pl-my gone-they 20 lbs little done- I have 'I have lost 20 lbs since my children left' 203. ruz-i Eahar mayl mi-dav-am day-a four mile pres-run-I 'I run four miles a day' The intransitive clauses representing state and process physical measurable case frames have obligatory objects functioning as complements filled by measure nouns. A measure noun may also fill the object-as-complement slot in transitive clauses (representing action-process case frames). 3.3.6.3. Identifier Identifier is a case which stands in equative relation- ships. In other words, identifier is a case which represents or identifies some other case on the surface structure (Sullivan 1980:309). In equative predications, something is either something else or becomes something else. Note the following examples: 204. masqasem model-e kuEek-e saxsiyat-e daijan bud Masqasem model-E small-E character-E Daijan was 'Masqasem was similar to Daijan in character' 145 205. mohamad-reza pahlavi axarin sah-e iran bud Mohammad—Reza Pahlavi last king-E Iran was 'Mohammad-Reza Pahlavi was the last king of Iran' 206. xahar-am belaxare modir-e madrese sod sister-my finally principal-E school became 'My sister finally became the principal of her school' Longacre uses the terms classifier/set as opposed to classi- fied/member to substitute for what we have, following Sullivan (1980), called identifier and patient, respectively. In other words, the identifier is a certain characteristic of the surface subject which is a deep patient in examples such as Jack is a president (Pike and Pike 1979:149) in English, and its Persian equivalent banisadr reise jomhur—e iran ast. However, we must note that identifier as a case is not commonly found with all predicates, being limited to equative and generic predications.14 3.3.7. Peripheral Cases (Roles) The twelve cases which were developed and explained in detail above are nuclear, and their roles are crucial to predication distinction in Persian. Causer (an argument in propositional calculus) and place (a non-essential argument in predication) as two examples of peripheral roles have already been dealt with (see sections 3.2.2 and 3.3.5). All peripheral roles are apparently within the category of non-essential roles since their existence is more pertinent at the sentence level .fi... 1). Si Pt 1... 146 than at the clause level. Time, manner, purpose, and so on are to be categorized as peripheral roles. Their de- tailed descriptions are beyond the scOpe of this work. 3.4. Devices for Predication Categorization What makes predicates be semantically close to one another are features attributed to each predicate (i.e., verb or verb phrase) in a case frame. We must note that features are not inherent specifications of cases but rather predicate features determine the nature of features attributed to case candidates. Thus the first step towards case analysis should be the identification of predicate features to enable us 1) to categorize similar or semanti- cally-tied verb types, and 2) to locate cases which charac- teristically and semantically match them (Longacre 1976:38). As already mentioned in 3.1.4., compound verbs in Persian, no matter whether they are analyzed into predicate-role relationships or not, are considered as phrasal verbs similar in action to the simple verbs. The number of features distinguishing verb types is not unlimited and could be summarized as follows: 3.4.1. Physical Physical is a feature of predicates in whose semantic domains a patient obligatorily occurs. It refers to phy- sical state, processes, or activities which are in one way or another related to physical or bodily activities. Verbs as 147 V 9 such as sekastan 'break', rixtan 'pour', sostan 'wash', xarab kardan 'demolish', harekat dadan 'move', dadan 'give', and so on have accompanying patients.15 In an example such as sir babr-r; zaxmi kard 'The lion wounded the tiger', babr 'tiger' is a patient which physically undergoes the action. Here, the nature of the predicate zaxmi kardan 'to wound' requires a patient. 3.4.2. Experiential Experiential is a feature attributed to predicates which require an experiencer in their immediate environments. The experiencer is an animate entity which emotionally, sensationally, or psychologically undergoes an event. Emotive, and sensation verbs, as well as speech verbs are to be attributed with an experiential feature. One who feels something, or is made to feel something, is the experiencer and not the patient. In an example such as sir hame-ye heyvanat-e jangal—ra mitarsan-ad 'The lion frightens all the animals in the forest', the nervous systems of all animals register a condition, fear in this case. 3.4.3. Ambient Ambient is a feature attributed to predicates which refer to environmental factors. In Persian hgyé, 'weather' as an item fills the subject slots in many examples. However, its occurrence in cases related to the environmental 148 conditions is optional, provided that a time margin is present in the sentence (see section 3.5.1 below). 3.4.4. Possession Possession or ownership is assigned to verbs which establish this relationship between one who possesses something and what is possessed. The possessor is the goal, whereas patient encodes what is possessed. 3.4.5. Directed Verbs which occur in desire and acquaintance or sensa- tion and speech case frames have categorically an obliga- V tory feature 'directed'. In the clause masih peyrovan-as- r3 dust darad 'Jesus loves his followers', the act of love is directed to the followers. 3.4.6. Intentional Verbs occurring in physical action-process case frames (see 3.5.7) may optionally have a feature 'intention' which, in turn, determines the nature of the performer (agent) of the action. In Persian, some verbs are distinguished on the surface with regard to whether this feature is present or absent in the predicates involved. Take kostan and be gatl resandan 'kill', both of which semantically refer to the occurrence of an event with both an agent and a patient as an object or undergoer. Both have instrumental as a feature. While the former optionally takes the feature 'intention' 149 the latter inherently possesses this feature. Thus, a sentence such as: 207. ali sag-r3 kost Ali dog-Ach killed 'Ali killed the dog' is ambiguous in the sense that Ali might have intentionally or unintentionally performed the action. He might have run over the dog and killed him accidentally. This ambiguity can be eliminated by the addition of a margin manner tasadofan 'accidentally' to the surface structure. Be qatl resandan is different in the sense that no ambiguity arises in this respect. Moreover, an instrument might optionally occur in the subject slot by tOpicalization if the verb kostan 'kill' rather than be qatl resandan is present. We might have clauses (208) and (209) below, but not (210) and (211). 208. qatel-e atlantéi faqat baEEe-ha—ra mi—kos-ad/ be qatl mi-resan-ad killer-E Atlanta only child-pl-Ach pres-kill- he to murder pres -comp1ete 'The Atlanta killer kills only children' 209. belaxare taryak pedar-am-ra kost finally opium father—my-Ach killed 'Finally, Opium killed my father' 210. *zahr soqrat-ra be qatl resand poison Socrates-Ach killed 'The poison killed Socrates' 150 211. *sang u-ra amdan kost rock he-Ach on purpose killed-it 'The rock killed him on purpose'16 3.4.7. Completable Predicates having a completable feature in their semantic configurations require ranges as their complements. The Persian verbs such as danestan 'to know', fahmidan 'to understand' amuxtan (yad gereftan) 'to learn', amuxtan (yad dadan) 'to teach', elqa kardan 'to impart', motale'e kardan 'to study', and so on, have this feature. Ranges may or may not surface, however. Speech verbs such as hekayat kardan 'to tell stories', goftan 'say', naql kardan 'quote' etc. have ranges rather than patients. In the following example, qesehaye ajib-o-qarib is a range: 212. masqasem qese-ha-ye ajib o qaribi baraye ma ta'rif mi-kard Masqasem stories strange and queer for we explain did 'Masqasem used to tell us strange stories' The predicate itself requires the range in order to be semantically complete. 3.4.8. Measurable A verb having measurable as a feature requires a role 'measure' as a quantitative completion to its predication. Rosd kardan 'grow', vazn dastan 'to weigh', kutah kardan 'shorten', etc. require measure as roles even though the arguments may not surface. 151 3.4.9. Locative Locative is the feature of those verbs in whose description a locative role is obligatorily present. The locative case may either overtly or covertly appear in the surface structure or be understood implicitly. Verbs such as gozastan 'put', nesandan 'to sit (transitive)', maxfi kardan 'to hide', mandan (eqamat kardan) 'to stay', andaxtan (gozastan) 'to put', etc. are differentiated from other verbs by this feature. 3.4.10. Motion In order for a path to occur in collocation with cer- tain verbs, the predicates corresponding with those verb types should have a feature motion in their underlying predications. Motion and activity should be distinguished. A motion is an activity, but not every activity is a motion. In the following example: 213. baradar-am yek atari baraye pesar-am xaride bud brother-my one Atari for son-my bought was-he 'My brother had bought my son an Atari' the predicate xaridan 'to buy' has an agent/path and a patient in its environment. Thus, it is characterized by the feature motion. There is a transfer of entities involved. 3.4.11. Instrumental Certain verbs inherently have the feature 'instrumental' in their underlying representation. Thus, javidan 'to chew' be Eh «Q Os .1)— 1: ‘ w». 152 implies that an instrument be present in the predication. However, this feature is so implicitly intermingled with the predicate that the case is totally covert, and its occurrence on the surface is redundant. 3.5. Case Frames17 In Chapter Two of this work, a detailed analysis of case frames of English was presented with the intention of providing a theoretical framework on which a tenable classification and case frame analysis of Persian could be based. A case frame was defined as a semantic configuration consisting of a verb with certain characteristics accompanied with a number of permissible roles (cases). The predicate characteristics are definable in terms of which groups of verbs occur in collocation with cases. A verb form might occur in two different rows but be differentiated by features and the accompanying roles. There are thirteen rows, con- sisting of sets of frames (see chart 2), the last of which is controversial. Whether to consider equative predications as equal with other predications has been questioned by Longacre (1976:95). However, there are certain reasons to believe that equational predications with identifier as their immediate case, should be allowed as a predication category, at least in Persian (see 3.5.13). 3.5.1. Group A: Ambient Case Frames Tabaian apparently rejects the notion that ambient predication occurs in Persian (1979:205-206). He states 0c is 51 no Co“ .. 5. “~33 153 that where English uses ambient verbs, Persian uses state verbs instead. In the corpus provided for the present work, very few examples of ambient predications were found. However, finding examples--particularly in everyday conversation-- to indicate the existence of case frames of this type is not a difficult task. State ambient case frames refer to environmental factors. In Persian, a highly general noun hgyé 'weather' optionally occurs in this case frame. Whether pgyé is a real patient is questionable. However, EEK; as a filler of the subject slot is frequently omitted in the state case frames, and no pronoun is substituted for it in less formal examples. Consider the following examples: 214. a. hava sard-e b. emruz sard-e c. sard-e weather cold-is today cold-is cold-is 'The weather/ 'The weather/ 'The weather/ It is cold' It is cold today' It is cold' 215. a. have garm-e b. emruz garm-e c. garm-e weather warm—is today warm-is wasrm—is 'The weather/ 'The weather/ 'The weather/ It's warm' It is warm today It's warm 216. a. have aftabi—st b. emruz aftabi-ye c. aftabi-ye weather sunny-is today sunny-is sunny-is 'It's sunny' 'Today is sunny' 'It's sunny' Apparently, the copula is the only possible verb in this case frame. Process ambient case frames refer to a change of climatic conditions. The following examples exemplify this case frame: 154 217. a. hava dare sard mise weather prog cold become 'It's getting cold' b. emruz dare sard mise Today prog cold become 'It's getting cold today' 218. a. hava dare barfi mise weather prog snowy become 'It looks like it's going to snow' b. emruz dare barfi mise Today prog snowy become 'It looks like it's going to snow today' The inchoative verb sodan 'become' signifies this process of change. The action-process ambient case frame is exemplified by the following rare example: 219. in deraxt-ha ba'es-e xonaki-ye (hava) sode—and this tree-pl cause-E cool-E (weather) become-they 'These trees have caused the weather to become cool' In the action ambient case frame, an assertion is made about what is happening in the environment. Here, the occurrence of hgyé makes the sentence unacceptable unless it is used as a nominal referring to an underlying source. Consider the following examples: 220. a. dare barf mi-yad prog snow pres-come b. *ere have barf mi-yad prog weather snow pres—come c. *hava dare barf mi—yad 'It's snowing' 155 221. a. dare barun mi-yad prog rain prog-come b. *dare hava barun mi-yad weather c. *hava dare barun mi-yad 'It's raining' 222. az hava dare barun mi-yad from sky prog rain prog-come 'It's raining' 223. az have dare barf mi-yad 'It's snowing' There is a tendency to equate hava in (222) and (223) with aseman sky etc. fall.18 as the supposed origin from which rain, snow, In accord with the above analysis the following case frames might be devised for state, process, action-process, and action ambient verbs, respectively. \ S-Ambient _ P-Ambient _ AP-Ambienak hava/O have/0 hava/¢ J /potent {A-Ambient 0 3.5.2. Group B: Ambient Experiential Case Frames Case frames in this group refer to the climatic condi- tions with the difference that, here, in contrast with the ambient case frames, the environmental factors register on an experiencer's nervous system. Khanlari (1973) mentions a group of so-called 'non-passing' verbs which, according to him, always exemplify the present conditions (320-21). 156 Some of his examples fit the examples presented in the following paragraphs. The state ambient experiential case frame semantically represents clauses as follows: 224. (man) garm-am-e 'I'm hot' I hot-me-is 225. (man) sard-am-e 'I'm cold' I cold-me-is 19 226. adam tu misigan narahat-e man in Michigan uncomfortable-is 'One feels uncomfortable in Michigan' (It's hot/cold) In these example, man 'I' and adam (lit. a human being) are experiencers whose nervous systems register the environmental factors. Khanlari states that " ...the meaning of this type of verb does not affect any nominal accompanying the verbs except the one being the logical subject [i.e., filling the subject slot] ..." (320). (My translation). The process ambient experiential case frame refers to the process of the environmental factors changing and an experiencer feeling that change. The following examples exemplify this case frame: 227. A: dare garm-am mise A: 'I'm getting hot' prog hot-me become B: kot-et-o dar ar B: 'Take your coat off' coat-your—Ach out bring 228. sarma sarma-m mise . . . . . I m shivering w1th cold cold-me become cold' 229. surat-am dare gol mi-yandaze 'My face is blushing' face-my prog flower prog-throw (I'm getting hot/cold) 157 230. dast- a—m dare ye tike yax mise hand-pl-my prog one piece ice becomes 'My hands are getting as cold as ice' The action-process ambient experiential case frame indicates an action being performed by an agent towards an end, with the effects being registered by an experiencer. In this case, the experiencer is coreferential with the agent. The surface reflexive 52$ 'self' indicates this coreferen- tiality. Examples are as follow: 231. dar-am xod-am-o tuye aftab garm mi-kon—am prog-I self-my-Ach in sun warm pres-do-I 'I am warming up myself in the sunshine' 232. ali xod-es-o pahlu-ye ates garm kard Ali self-his-Ach beside-E fire warm did-he 'Ali warmed himself up by the fire' We should note that the more general nominal hava cannot act as an agent characterized with the feature potent. Thus examples such as: 233. *(?) have dare badan-am-ra garm mi—kon-e weather prog body-my-Ach warm pres-do-it 'The weather is warming up my body' are not acceptable. Examples of the action ambient experiential case frame are odd and unacceptable to many native speakers of Persian. Examples such as: 234. (?) bar man barid 'It rained on me' on I rained with no antecedent are odd. 158 The following case frames may be devised to indicate role-predicate relationship in this group: J’ 'L AP-Ambient S-Ambient. P-Ambient E E A/ E [Exper Exper J Exper 3.5.3. Group C: Experiential Case Frames The state experiential case frame includes verbs which signify emotive states in which an experiencer's nervous system registers the conditions. An instrumental feature can optionally occur with the predicate. As far as my data show, these predicates surface in the so-called compound verbs characterized by verbals (Khanlari's 'hamkerd') dastan 'have', budan 'be'. The state experiential case frame under- lies the following clauses: 235. ali az zan-as haras dare Ali from wife-his fear has 'Ali is scared of his wife' 236. emruz maryam sangul-e today Maryam happy-is 'Maryam is happy today' 237. daijan saxt barasofte bud Daijan hard upset was 'Daijan was restless' 238. daijan az adam-e movafaqiyat-e xod saxt qamgin bud Daijan from lack-E success-E self hard sad was 'Daijan was very sad because Of his lack of success' az zan-as 'of his wife' in (235), and az adame movafaqiyat 'his lack of success' in (238) are instruments (stimuli). 159 The process experiential case frame involves an ex- periencer who senses a psychological change. The Optional instrumental feature may correlate with the verb sodan 'become' in the surface structure of Persian (Tabaian 1979: 205), 'compound verbs' with verbals sodan 'become', kardan 'do', and other simple verbs as well. Consider the following examples: 239. baEce-ha az sibil-e yaru 33 xord-and child-pl from moustache-E fellow shocked-they 'The children were shocked when they saw the fellow's moustache' 240. dustalixan pasiman sod Dustalixan regretful became 'Dustalixan felt regret' 241. m3 az in nane-as tarsid-im we from this mother-his feared-we 'I got scared of his mother'20 242. man kam kam tars-o-vahsat-am rixt I little little fear and fear-my poured 'Gradually I lost my fear' The fellow's moustache in (239), from his mother in (241) are instruments (see 3.3.4) which optionally occur in this case frame. On the other hand, baEEeha 'children', Dustalixan, and man 'I' are experiencers whose nervous systems register the psychological conditions. In the affective or action-process experiential case frame, an agent intentionally acts in such a way that his performance affects an experiencer. The predicates may be 160 Optionally characterized with a verbal instrumental feature which surfaces as an Optional instrumental case. The action- process verb is predominantely realized as kardan 'do'. However, the occurrence of other simple and compound verbs is possible. The action-process experiential case frame is exemplified in: 243. u mara be be qas zadan-e xod zahre tarak kard he me with to faint hit—E self gall-bladder burst did 'He shocked me by making himself faint' 244. mo'eze-ye kesis u-ra Bram kard preaching-E priest he-Ach calm did 'The priest's sermon cooled him off' 245. zan-as hamise be u sarkoft mi-zan-ad wife-his always to him reproach pres—hit—she 'His wife always reproaches him bitterly' In (244), the agent is characterized as potent (see 3.2.3). In the action experiential case frame, the objects are experiencers rather than patients since they do not phy- sically undergo changes. This case frame underlies the following examples: 246. ali doxtat-a§-t§ b3 lab—e Eatb—eé busid21 Ali daughter—his-Ach with lip-E greasy-his kissed 'Ali kissed his daughter with his greasy lips' 247. ali b3 pZ-§ zad be sag-e22 Ali with foot-his hit to dog-Def 'Ali kicked the dog (with his foot)‘ Here, the verbs are more physical, in contrast with the verbs in the affective case frames. However, the recipients 161 of the actions experience the action through their nervous systems. The instrumental roles accompanying the verbs such as busidan 'to kiss', lagad zadan 'to kick', navazes kardan 'to pat', and so on rarely surface except when an elaboration of them is demanded. In some examples, a range also occurs. The case frames in this group can be formulized as follow: { r \ ‘ 0 S-Exper p-Exper AP-Exper A-Exper E E (inst) A E (inst) A E (inst) (inst) intention; intention J 3.5.4. Group C': Factual Knowledge Case Frames The predicates in this group are all characterized by the feature/completable/, since the obligatory nominals in juxtaposition with the predicate, aside from the experiencers, are ranges. Verbs such as danestan 'know', fahmidan 'know' (lit. understand), ta'lim dadan 'teach', and so on fall within this category. The state experiential completable case frame underlies the following examples in which the underlined nominals (nouns, noun phrases) correlate with the occurrence of range as a role which surfaces as an object complement: 248. xoda xodes hame Eiz-ra mi-dan-ad God himself all thing pres-know-He 'God knows everything' 249. mi-dun-i ke tars dar qamus-e man vojud na—darad pres-know-you that fear in dictionary-E my exist neg-have 'You know that I don't know the meaning of fear' 162 250. m3 Eiz-e dorosti ne-mi—dan-im we thing-E true neg-pres-know-we 'We don't know the truth' 251. m3 ne-mi-fahm-im we neg-pres-understand-we 'We don't know' (lit. we don't understand (it)) 252. mo'alem-e ma xeyli riyazi mi—dun-e teacher-E we much math pres-know-he 'Our teacher knows a lot of mathematics' In (251), the object correlating with the underlying range is covert but could be realized on the surface by general terms such as Egg; 'something', matlabi 'a matter, etc. The process experiential completable case frame refers to the predicates with an experiencer which undergoes a prog- ression. Verbs such as amuxtan 'to learn' (yad gereftan), faramus kardan 'forget', fahmidan 'to learn' (lit. under- stand) occur in this case frame illustrated by the following examples: 253. hasan dare kam kam mozu-ra mi-fahme Hassan prog little little matter-Ach pres- understand 'Hassan is gradually understanding the matter' 254. aqa, ma mozu-ra yademun raft Sir, we matter-Ach our memory went 'Sir, we have forgotten the matter' 255. baEEe-ha dar arze §1§ mah ingilisi yad gereft—and child-pl in within six month English memory took- they 'The children learned English in six months' 163 In (253), fahmidan is semantically similar to yad gereftan 'learn'. In the action-process experiential completable case frame, an agent is the one who intentionally makes efforts to impart his/her knowledge to an animate entity who experi— ences the knowledge imparted. Consider the following examples: 256: asdolamirza asrar-e esq-ra be man yad dad Asdolamirza secrets-E love-Ach to me taught 'Asdolamirza taught me the secrets of love' 257. belaxare metlab-ra be u tafhim kard—am finally matter-Ach to him understand(tr) did—I 'Finally, I made him understand the matter' 258. aqabozorg savad-ra tuye kale—ye ma foru kard Aqabozorg knowledge-Ach to head-E us imparted 'Aqabozorg imparted the knowledge to us' 259. aqa-ye ahmadi dars mi-dah-ad Mr. -E Ahmadi lesson pres-give-he 'Mr. Ahmadi teaches' In (259), the experiencer is covert, but it is understood from the context. Note that amuxtan can surface in a trans— itive as well as an intransitive clause. However, in modern Persian, there is a tendency to substitute yad dadan 'teach' and yad gereftan 'learn' for the so-called surface transitive and intransitive clauses, respectively. In action experiential completable case frames, an agent coreferential with the experiencer occurs. Thus, the activity carried out by the agent impinges on his/her own 164 nervous system. Verbs such as motal'e kardan 'to study', az bar kardan 'to recite', and so on are in this category. Note the following examples of this case frame: 260. doxtar-am jadval-e zarb-ra hefz kard daughter-my multiplication table-Ach recited 'My daughter recited the multiplication table' 261. hasan ketab-ha-ye zabansenasi motale'e mi-kon-ad Hassan book-pl-E linguistics study pres-do—he 'Hassan studies linguistic books' 262. hasan motale'e mi-kon-ad Hassan study pres-do-he 'Hassan studies' A contrast between (261) and (262) shows that in (262) the object has been suppressed and the clause has an intransi— tive form on the surface. The formulizations for case frames in this group are as follow: I I a / S-Exper P-Exper AP-Exper A-Exper ER ER A E R A/E R complet complet complet complet / 3.3.5. Group D: Experiential Directed Case Frames In these case frames, an experiencer and a goal are nuclear roles in juxtaposition to the predicate. The verbs are characterized by the feature/directed/ in the sense that there is a motion-like flow from the experiencer's desire and cognition towards what is actually desired. Verbs such as xastan 'want', dust dastan 'like/love'. setayes kardan 165 'praise', senaxtan 'to know', and so on occur in this group of case frames. Goals occurring in these frames may be animate or inanimate. The state experiential directed case frame underlies examples such as the following: 263. baradar-e man elizabet taylor-ra dust dar-ad brother-E I Elizabeth Taylor-Ach love 'My brother loves Elizabeth Taylor' 264. daijan aseq-e napele'on bud Daijan lover-E Napeleon was 'Daijan loved Napeleon' 265. man axlaq-e soma-ra mi-dan-am I character-L you-Ach pres-know-I 'I know your character' 266. in jens—e xabis—ra man Sal-ha-st mi-senas-am this material-E dirty-Ach I year-pl-is pres-know-I 'I have known this dirty rat for many years' 267. maryam sohar mi-xah-ad Maryam husband pres-want-she 'Maryam wants a husband'23 In example (263), Elizabeth Taylor is the goal towards whom my brother's love is directed. However, if the goal is an animate entity which is aware of the experiencer's affection, it is both goal and experiencer simultaneously. In example (265), danestan is a surface realization of senaxtan as it is used in example (266). However, the concept KNOW in 1 Group C' above and KNOW2 in this section are normally realized differently on the surface. Longacre's example (59): 166 How can we know1 that we known God? has the Persian equiva- 1 L lent clause: 268. cegune be-dan-im ke xoda-ra mi-sanas-im How S-know -we that God-Ach pres-know -we 1 2 in which danestan (state experiential completable case frame) and senaxtan (state experiential directed case frame) sur- face differently. The process experiential case frame has an experiencer and a goal, with the difference being that the experiencer, instead of being in a certain condition, undergoes some emotional change. Note the following examples: 269. doxtar az tah-e qalb be aspiran aseq sode bud girl from bottom-E heart to Aspiran lover become was 'The girl had deeply fallen in love with ASpiran' 270. maryam b3 ali asna sod Maryam with Ali familiar became 'Maryam got acquainted with Ali' The action-process experiential case frame underlies the following clauses: 271. maryam-r3 b3 ali asna kard-am Maryam-Ach with Ali familiar did-I 'I introduced Maryam to Ali' 272. zan-g hamsaye maryam-r5 be nefrat az sohar-as va dast wife-E neighbor Maryam-Ach to hatred from husband— her forced 'The neighbor's wife made Maryam hate her own husband' 167 Note that Ben; kardan 'to introduce' can be derived from Etna sodan 'to be acquainted with', but it is not possible to derive *aseq kardan from aseq sodan 'to fall in love'. In action experiential directed case frame, an agent performs an action whose effects are either directed towards; or experienced by an animate entity. Examples are as follows: 273. hasan maryam-r3 taqbih kard Hassan Maryam-Ach contempt did 'Hassan held Maryam in contempt' 274. bani sadr xanom-e taEer-ra bexatere marg-e baby sandz sarzanes kard Bani Sadr lady-E Thacher-Ach for death-E Bobby Sands reproach did-he 'Bani Sadr blamed Mrs. Thatcher for Bobby Sands' death' 275. osvald-ra be qatl-e reis jomhur motahem kard-and Oswald-Ach to murder-E chief republic condemnation did-they 'Oswald was condemned for the president's murder' All clauses of this type have two non-peripheral cases. The animate entity filling the direction terminal may be goal or goal/experiencer. The formulizations of the Persian case frames in this group are as follows: I \ S-Exper P-Exper AP-Exper A-Exper EG EG A E G A [E/E] directe directed directed directed 3.5.6. Group D': Experiential Completable Directed Case Frames In these case frames, a reverse situation occurs, that is, a sound wave, a light wave, or an odor emanating from the source is directed towards an experiencer and is sensed 168 by him/her. The features 'direction' in group D and D' are slightly different in the sense that, in the case frames in group D', a more physical activity is carried out. That may be the reason why a state experiential completable directed case frame does not occur. The process experiential completable directed case frame is exemplified as follows: 276. man yek aseq na-dide bud-am I one lover neg-seen was-I 'I had never seen a person in love' 277. man dard o qose-ye soma-ra mi-fahm-am I pain and grief-Ach you-Ach pres-understand-I 'I feel/know your frief and pain' 278. dar seda-ye madar-am negarani-ye u-ra hes mi-kard-am in voice-E mother-my anxiety-E she-Ach feeling pres-did-I 'I could sense my mother's anxiety in her voice' 279. az aseq sodan-am lazat mi-bord-am from lover become—my enjoyment pres-took—I 'I was enjoying my sense of love' 280. az yek nafar az famil sanide ke b3 zan-i rabete dar-am from one person from family heard that with woman-a relation have-I 'She has heard from a member of the family that I have a secret affair with a lady' 281. baEEe-ha seda-ye parande-r3 sanid-and child-p1 sound-E bird-Ach heard-they 'The children heard the birds sing' In (277), fahmidan is different from its homonym in group C', since, in this context, it must be interpreted as ehsas kardan 169 'to feel'. In (281), the source and range should obliga- torily be present in the deep and surface structures since (282), as a corresponding clause to (281) is unacceptable. 282. *baEEe-ha parande sanid-and child-p1 bird heard-they The action-process completable directed case frame most often refers to verbs of speech actions. This case frame underlies the following clauses in Persian: 283. asdolamirza dar hamam avaz mi-xand Asdolamirza in bathroom song prog-sang 'Asdolamirza was singing in the bathroom' 284. mi-goft-and Esq hame-ye dard ast IND-said-they love all-E pain is 'It was said that love is nothing but pain' 285. u avaqeb-e vahsatnak-e esq-ra baray-am sarh dad he consequences-E fearful-E love-Ach for-me explanation gave 'He told me of the consequences of love' 286. faramarz naqasi-ro be man nesan dad Faramarz painting-Ach to me showed 'Framarz showed me the painting' In examples (283) and (284), the experiencer is partially covert. In other words, an experiencer might optionally be chosen on the surface. In action experiential completable directed case frames, the experiencer is active in the sense that it exerts energy to comprehend, to listen, to watch, etc. Thus, in this case frame, the verbs gus dadan 'to listen', tamasa kardan 170 'to watch', Eesm duxtan 'to gaze', and so on are used with an agent/experiencer. Examples are as follows: 287. aqaaan be noqte-ye durdasti xire éode bud Aqajan to point-E far away gaze become was 'Aqajan was gazing at a point a long distance away' 288. Eesm az dahan-e daijan bar ne-mi—dast eye from mouth-E Daijan away neg—IND-has-he 'He was gazing at Aqajan' 289. masqasem cesm be qeyEi-ye baqbani duxte bud Masqasem eye to scissors-E gardening sewn was 'Masqasem was looking at the garden shears' The conflated formulizations of case frames in group D' is as follows: ‘v P-Exper AP-Exper A-Exper complet ER(S) complet A/S E R complet A/E R(S) pdirected directed} directed k 3.5.7. Group E: Physical Case Frames As the chart of verb distribution shows (see p.94), the bulk of verbs used in the corpus fall into this category. There is one significant difference between the verbs in this group and those below and those discussed in the sections above. Some verbs in state, process, and action-process are derived in progression by the substitution of sodan 'become', and kardan 'do' in process and action-process, respectively, for the more general verbal budan 'be' in state verbs. This rule, however, cannot be generalized, as shown in the examples below: 290. But 291. 171 a. xosk budan 'be dry' xosk sodan xosk kardan b. eftetah budan'be Open' eftetah sodan eftetah kardan c. pir budan 'be old' pir sodan pir kardan d. xarab budan'be ruined' xarab sodan xarab kardan a. xab budan 'be asleep' *xab sodan24 xab kardan b. zan budan 'be a woman' *zan sodan25 *zan kardan26 c. band budan 'be obstructed' band amadan band avardan The state physical case frame contains a patient and a predicate; in turn, the predicate is composed of an adjective plus the verb budan 'be' or dastan 'have'. The state physical case frame is exemplified by the following clauses: 292. 293. 294. 295. in Befaq O dosmani dar xanevade-ye ma qabel-e davam nist this discord and enmity in family-E our capable-E continuation isn't 'The animosity in this family cannot continue' bayad qavi baa-am must/should strong be-I 'I should be strong (physically)' dustali zende ast Dustali alive is 'Dustali is alive' daste-ye anu sekaste ast handle-E knife broken is 'The handle of the knife is broken' It is interesting that budan and dastan are mutually exclu- sive in this case frame if the attributive adjective is 172 accompanied with gabel ' able'. e.g. 296. davam dastan 'to last' qabele davam budan/*dastan baqa dastan 'to continue qabele baqa budan/*dastan to exist/ survive' The process physical case frame involves a patient which undergoes a physical action. An optional instrumental case may occur with the predicate. Consider the following examples: 297. 298. 299. 300. 301. 302. In (302), ezterab-e puri askar sod anxiety-E Puri appear became 'Puri's anxiety came to light' seda-yam band amad voice-my blocked came 'I couldn't speak' ketab-ha-ye aseqane kam 33p sode bud book-pl-E amorous little print become was 'Very few books about love had been published' estentaq edame yaft interrogation continuation found 'The interrogation continued' samavar mi-jusid samovar prog-boiled 'The samovar was boiling' tamam-e badan-am az tars mi-larzid all-E body-my from fear prog-shivered 'I was trembling' az tars is an instrumental stimulus. In (301), the patient (i.e., ER 'water') is covert and the container occurs in the subject slot. 173 In action-process physical case frames, as in other action-process case frames, predication has both an accompany- ing agent which initiates a physical action and a patient which undergoes the action. However, despite the stability of the role of the patient, the agent may be either an instigator, an initiator performer or simply an unintentional figure who happens to represent an argument in relation to the predicate. The following examples should clarify: 303. ingilisi-ha u-ra kost—and English-p1 him killed—they 'The English people/troops killed him' 304. savaran-as mara zir—e dast o pa-ye asp-ha 1eh mi-kard-and horse-riders-his me under-E hand & foot-E horse- pl smashed 'I was being trampled under the hooves of his soldiers horses' 305. to madar dar in Ear 531 mara dah $31 pir kard-i you mother in this four year me ten year old did-you 'You aged me for ten years in your four years of absence' 306. masqasem rise-ye deraxt-ra qat' karde bud Masqasem root—E tree-Ach cut done was 'Masqasem had cut the root of the tree' 307. be golule se ta az dozd—ha-ra xaband ruye zamin with bullet three from thief-pl-pl-Ach sleep (tr) on ground 'He shot three thieves with one bullet' 308. m3 in divar-ha-ra xord—o-xamir mi-kon—im we this wall~pl-Ach smashed-&-dough pres-do-we 'We'll demolish these walls completely' 174 The question is how to differentiate between an action being performed intentionally and an action with an unintentional performer. The feature intentionality versus unintentional- ity should be attributed to the predicate which determines the roles. For instance, in (304), (305), (303) and (306), the clauses have two interpretations, whereas the agents in (307), and (308) act intentionally. This interpretation might be partially related to the nature of the predicates as well as to the existence of some kind of instrument in the predication, whether it appears overtly or covertly. A further problem is the agent in (305). Is t 'you' an instrument or an agent? The former interpretation is more probable since pp you is not particularly involved in the action but is rather an instrument to that action. The action verbs in the action (physical) case frame are more pertinent to the nature of predication than what the action performs on. In other words, the effects of the action are not on the surface object as they have been ex- emplified in the action-process case frames. Note the following examples: 309. rajabali bi- xabar az man zan gereft Rajabali without-news from I wife took-he 'Rajabali (had) married without letting me know' 310. nafas-i kesid-am breath-a pulled-I 'I inhaled' 175 311. some ne-mi-tavan-i zan be-bar—i you neg-pres-can-you wife/woman S-take-you 'You cannot marry' 312. pedar-e daijan haft emarat saxte bud father-E Daijan seven building built was 'Daijan's father had built seven buildings' 313. sobhane-ra xord-am breakfast-Ach ate-I 'I ate (the) breakfast' 314. dustali xare ba most zad ruye miz Dustali donkey with fist hit on table 'Dustali the fool knocked on the table with his fist' Examples (309) and (311) have integral verb forms. In other 2 words, zan bordan 'to marry' and zan gereftan'to marry' are considered as having a single meaning. In (312), and (313), the surface objects are ranges, whereas in (314), the object is a goal towards which an action is directed with no physical or spatial changes involved. The formulization of this group is as follows: I W V I ! S-Phys' P-Phys AP-Phys A- w 1 P P AP A R 9 Lintention ' G L , J \ 3.5.8. Group F: Measurable Case Frames There were few examples to exemplify this group in the data. However, it is not hard to find other examples for the following case frames. A state physical measurable case frame consists of a predicate and at least two obligatory cases: patient and 176 measure. The following clauses underlie this case frame: 315. fasele—ye ma sad qadam bud distance-E we 100 step was 'The distance between us was about 100 steps' 316. nazdik-e do metr qad dast close-E two meter height had 'He was about two meters tall' The underlined nominals (noun phrases) are measure. The process measurable case frame has a case frame similar to the one for the state case frame, with the exception that the predicate has process as its feature. The feature physical in the predicate requires a patient case. Examples follow: 317. baby sast pond laqar sode ast Bobby 60 lbs thin become is 'Bobby has lost sixty pounds' 318. qeymat-e tala dar bazar dah dar sad tanazol kard price-E gold in market 10 in hundred decline did 'Gold devaluated 10% in the market' 319. sath-e ab-e rudxane yek metr bala amad level-E water-E river one meter up came 'The river flooded by one meter' The action-process measurable case frame has an agent which initiates the action, a patient which undergoes the action, and a measure case. The measure role can either surface or remain covert. This case frame underlies the following clauses: 177 320. salvar-am-ra dah santimetr kotah kard—am pants-my-Ach 10 centimeter short did-I 'I shortened my pants ten centimeters' 321. parce-ra yek metr borid-am cloth—Ach one meter cut-I 'I cut the cloth off one meter' 322. jade-r3 dah metr ariz kard-and road-Ach 10 meter wide did-they 'They widened the road ten meters' The action measurable case frame has no patient, involving rather only an agent and a nuclear measure. Here, too, the measure may not occur on the surface, but rather be implied on the basis of the deep structure. Consider the following: 323. dar ebteda-ye jang eraqi-ha Eehel kilometr pisravi kard-and in beginning-E war Iraqi-pl 4O kilometer advance did-they 'At the beginning, the Iraqis advanced 40 kilometers' 324. do metr pares-e ertefa kard two meter jumping-E height did 'He highjumped two meters' The following is the formulization of these case frames: } \ ’ \ S-Phys 8 P-Phys AP-Phys ‘A- PM PM PM AM Lmeasurable imeasurable measurabl measurable 3.5.9. Group G: Locative Case Frames All predicates underlying verbs in this group must have a locative feature in their predications and a covert or overt locative case. Note that locatives are different from 178 goals, paths, and patients (see 3.3.1.2 and 3.3.3.2) in that no action is directed to or away from the location. In the state locative case frame, there are a patient and a locative in juxtaposition to a predicate. The following exemplify this case frame: 325. 326. 327. in sa'at dar xane-ye m3 bud this watch in house-E we was 'This watch was in our house' dustali dar Elam—e ro'b o vahsat bud Dustali in world-E fear and fear was 'Dustali was in his world of fear' araq be pisani-ye daijan nesaste bud sweat to forehead-E Daijan sat was 'Daijan's forehead was wet with sweat' The action-process locative case frame involves a feature/locative/ ascribed to the predicate and an agent, a patient, and a nuclear locative case in the predication. This case frame underlies the following examples: 328. 329. 330. masqasem bil-r3 ruye sun-as gozast Masqasem spade-Ach on shoulder-his put 'Masqasem put the spade on his shoulder' xodam mi-zar-am-et tuye tabut myself pres-put-I-you in coffin 'I'll put you in a coffin' dobare azizosaltane-r3 sar-e ja-yas nesand-and agian Azizosaltane-Ach head-E place-her sit (tr)-they 'Azizosaltane was forced to sit down' 179 331. parsal tuye baqte-ye kelisa kalam kast-am last year in garden-E church cabbage planted-I 'I planted cabbage in the Church's garden last year' 332. mars andaxt ruye dug-as me threw on shoulder-his 'He put me on his shoulder' The case frame in example (332) is different from a similar predication with preposition 23 instead of ppyg. In the former, the nominal would be a goal rather than a locative.29 The action locative case frame involves predicates with an agent which fills the surface subject slot and an obligatory locative case. Consider the following examples of this case frame: 333. man xodra labels—ye deraxt-ha maxfi kard-am I myself among-E trees hidden did-I 'I hid among the trees' 334. some dar moqabel-e yek adam-e mosamam qarar dar-id you in front-E one man-E determined situated have you 'You are standing in front of a serious person' 335. dar otaq-e entezar mand-am in room-E waiting stayed-I 'I stayed in the waiting room' Notice that if the animate agents £223 in (334), and (BEE) 'I' in (335) are substituted for by inanimate nominals, the case frames will be state locative rather than action locative case frames. The reason might be that for animate entities to stand or to remain in a location requires exerting energy 180 (Longacrez73). The following formulizations can be de- vised for the Persian case frames in this group: S-Phys AP-Phys A- PL APL AL locative locative locative \ 3.5.10. Group G': Physical Motion Case Frame In this group, predicates are characterized with the feature/motionl, which in theory cannot collocate with locative cases. These verbs imply that actions are carried out towards a goal or through a path. Source is optional in these case frames. The process physical motion case frame involves a predicate characterized by the feature/motion/ and at least one patient. Optional occurrences of path, source, and goal are also possible. This case frame underlies the following clauses: 336. masqasem az ruye taxt be zamin oftad Masqasem from on bed to ground/floor fell 'Masqasem fell off the bed' 337. moqe'-e xab hafttir az-as joda ne-mi-sod time-E sleep revolver from-him separate neg-IND- became 'He always had his gun with him even when sleeping' 338. mgsl-e baran golule az in taraf 0 an taraf mi— barid like-E rain bullet from this side and that side rained 'Bullets rained from everywhere' 181 339. sini az dast-e u vel sod tray from hand-E him released became 'The tray fell from his hand' 340. asar o ala'em az surat-as mahv sode bud traces and signs from face-his disappear had become 'All signs had vanished from his face' In example (336), the predicate indicates that Masqasem undergoes the action, but there is no agent to initiate the action. The goal and source are overt, although they could also have been covert. In (337), the action and the under- goer are important,a1though one can postulate an agent to move the patient around. In (338), a sentence margin simile occurs, and there is also a source from which the action originates. In (339), dast—e u 'his hand' is the source, but the goal is covert; it could be overtly signalled by a prepositional phrase such as be zamin 'to the floor', ruye zamin 'to the floor' and the nominal phrase kafe otéq 'the floor'. In (340), the patient is a more concrete sign which has disappeared, thus it has undergone a change of location. Surat-as 'his face' is the source. The action-process case frame has an Obligatory agent and an obligatory patient in juxtaposition to the predicate. Path and goal are optional. The following examples illus- trate this case frame and the type of verbs which can occur in this framework: 182 341. caqu-ra besuye man partab kard knife-Ach towards me threw did 'He threw the knife towards me' 342. leyli dast-as—ra az dast—e man birun kesid Leyli hand-his-Ach from hand—E I out pulled 'Leyli took her hand from mine' 343. puri mara betarafe surax hol dad Puri me towards hold pushed 'Puri pushed me towards the hole' 344. qasoq-ra az panjere-ye otaq-e xod be daxel-e aspazxane-ye hamsaye andaxt-am spoon-Ach from window-E room-E my to inside-E kitchen-E neighbor threw-I 'I threw the spoon into the neighbor's kitchen through the window of my room' In all the examples above, predicates show motion towards, away, through, or out of a location. The action physical motion case frame involves an agent/patient3O which fills the subject slot on the surface, and optional goal, path, and source. Note the following examples: 345. ruye taxt-am az in dande be an dande misodam on bed my from this rib to that rib I was becoming 'I was tossing and turning in my bed' 346. az postebam betarafe daxele-e xane sarazir sod from roof towards inside-E house went 'He left the roof and went down to his house' 347. xodam-r3 resand-am be qole—ye taxtesang myself- reach(tv) to tOp-E rock 'I pulled myself to the tOp of the rock' 183 Despite other examples, in (345) and (346), the verb forms morphologically correlate with the coreferentiality of agent and patient. If the agent and patient are two diff— erent entities, then (348) is used instead of (345). 348. ruye taxt az in dande be an dande-am mikardand 'I was tossed and turned in my bed' These case frames are formulized as follows: \ P—phys S AP-Phys P ( path ) A/SP (path) (goal) motion goal motion A-Phys A/P (path) (goal) motion I 3.5.11. Group H: Physical Possession Directed Case Frames Two major differences differentiate groups H and H' from those discussed in previous sections: a) path, goal, and source are animate entities which possess, own, give, receive, or hand over their prOperties; and b) a further feature/possession/ is attributed to the case frames (see section 3.3.1 and 3.3.3 for the identification of path, goal, client, and source). Furthermore, group H has a feature/directed] in its predicates. The state physical possession directed case frame in- volves a patient (what is possessed), and a goal (one who possesses something). The following clauses underlie this case frame: 349. 350. 351. 184 hala digar yek otomobil-e no dar-im now one car-E new have-we 'Now, we have a new car' anha sarnevest-e qamangizi dast-and they fate-E sad had-they 'All met tragic fates' zahmatkesan-e iran zendegi-ye sagi dar-and workers-E Iran life -E dog have—they 'The workers in Iran have miserable lives' The process physical motion directed case frame has a process verb, a patient, and a goal. Very few examples of this case frame were observed. Consider the following: 352. 353. taxti medal-e tala gereft Taxti medal-E gold took 'Taxti obtained (a) gold medal' puri az danesgah lisans gereft Prui from university B.A. took 'Puri got his B.A. from the university' The action-process physical motion directed case frame is similar to the process possession case frame with the exception that an agent is obligatorily present in the predication. Note the following examples: 354. 355. nim-i az bastani-ye xod-r3 be leyli dad-am half-a from ice-cream-E self-Ach to Leyli gave-I 'I gave Leyli half of my ice-cream' baradar-am mah-i sisad dolar be man mi-deh-ad brother-my month-a 300 dollars to me pres-gives-he 'My brother gives me 300 dollars each month' 185 In the examples above, man 'I' [covert] in (354), and baradaram 'my brother' in (355) are agents/sources. The action physical possession directed case frame is exemplified by the following example: 356. pesar-e man baraye sargarmi kebrit jam' mi-kon-ad son-E my for amusement matches collects 'My son collects matches as a hobby' 3.5.12. Group H': Physical Possession Motion Case Frames The case frames in this group are similar to those in Group H with two essential differences: a) the feature/ directed/ in group H is replaced by the feature/motion/ in group H', and b) a path (transitory owner) obligatorily exists in the case frames. Not many examples were found in my corpus, however. The following clauses exemplify the state and process case frames in this group: 357. hedye-i baraye nozad-e some dar-am gift-a for baby-E your have-I 'I have a gift for your baby' 358. xodavand baraye hame-ye bandean-as ruzi dar-ad God for all-E creatures-his daily-bread has-He 'God gives daily bread to all his creatures' 359. yek kif—e Earmi baraye doxtar-am peyda kard-am one bag-E leather for daughter-my found-I 'I found a leather bag for my daughter' 360. yek bilit-e ta'atr baraye xahar-at be dast avard-am one ticket-E theatre for sister-your to hand brought-I 'I got a ticket for your sister' 186 (357) and (358) are state, and (359) and (360) are process case frames, respectively. In action physical motion possession case frames, an agent/path combination at the predication level is more common than an agent/source predicate relationship. Thus we find examples such as: 361. man yek gardanband-e tala baraye zan-am xarid-am I one necklace-E gold for wife-my bought-I 'I bought a gold necklace for my wife' The conflated formulization for case frames in groups H and H' are as follow: S-Physical directed. E G 'P-Physical directed. F G possession motion [P Path G possession motion. E Path G AP-Physical directed [A/S P G A- Physical directed [A/G P possession motion I} P PathG possession motion [h/Path PG 3.5.13. Group I: Equative Case Frames Longacre considers equation as a statement of set member- ship (1976:273). In other words, in a sentence such as Mr. Reagan is a president, a president is a set and the statement is in general an affirmation of the fact that Mr. Reagan is a member of the set. Longacre explicitly states that equation should be considered outside of the case frames and predication scheme (95). Pike, however, considers equative as a clause type with subject functioning as an item and the nominal as a complement characteristic of subject (1977:36; 46). According to Pike, in John is chairman, 187 ippp is the subject of the clause with a function of item, and chairman is a complement characteristic of subject. Pike uses this clause type for examples in which something is something else, or something becomes something else. However, many examples presented by Pike (46) can be cate- gorized into different rows in Longacre's (1976) chart. Whereas the surface structures are similar, the deep structure (case frames) are not. Some of Pike's examples will be used later to justify why an equation group has been postulated for Persian. Identifications of constituents are those of Pike. Identification references in parentheses are mine. S—It co-COS31 362. John became a man (John: experiencer {P-Expe%> E S-It co-COS I ‘ P 363. John is tall (John: patient tS-Phys [ — s / S-It co-COS 364. The milk turned sour (milk: patient P-Phys}. P k _. In the classification of Persian verbs, we have already come across many persian clauses similar to the equative types Pike and Pike present in their 1977 work. What we classify as equative predications in Persian will be clause types such as Np is/resembles NP, NP becomes NP, and X is turning NP into NP. As mentioned above, clause types such as NP be adjective, and NP becomes adjective have been taken care of so far by various parameters in the Persian predication identification. 188 The inclusion of these case frames in the predicate calculus is based on the fact that a) as do other predica- tions, the equatives need complements (labelled here as 'identifier'), b) they can have progressive constructions, c) the identifier, as with other cases, can be expanded, and d) verbs are not restricted to the copula. The state equative case frame underlies the following clauses: 365. masqasem model-e kuEek-e saxsiyat-e daijan bud Masqasem model-E small-E personality-E Daijan was 'Masqasem was similar to Daijan in character' 366. arafat rahbar-e Eerik-ha-ye felestin ast Arafat leader-E guerrilla-pl-E Palestine is 'Arafat is the leader of Palastanian guerrillas' Many metaphoric expressions might be analyzable in this framework. e.g.: 189 Onsmum— usow .c '|I"'I"II|""'II'-IIIJ I-|l|-|||"""|l-"l I c an ;uaa\< couuoa who; es\&_eaeutt.e asaeu< :O«OOI . .uaa m AcOA_ O a\< asteu< ---: ......... aunutuut O>_ spool _ < |< rug 0.35235:- e < s -< ewuouwup Amy: xx eenaeou axm < hag-J m\c d cw<.eaeeee.= a———J: :EOO "us—LEOO Owuzcm "m. .eaea e essaualfi u=O«sa< ” assuuos " u :3: quooo~ ; ea < case a< 1 : wma: no <— ||||||| II'II -----.‘4' \ u: uuuuuuuuuuuuu utnnnuuuul z=_s;< mwmucxmlzc~ku< :o_uoo ” mmOucC:|:o_OOm " LE< messes: " a; OOan u stunt: uuuu nanuuuuustn 4 .c_ea es:e-e_ 3. run: a; :o_uoa woos u eeweaeeerav asaa-eh - ........... a----amwwwwa. A sex; was m asaela rile-ll lllllll [8| lllll IIOII : a; o—sstzacOs a; a wa:gl paqusuv lac e e e._eeee axmla I'll-Ill-" ..... II |||||||| J c w canons—t E z m uOmeIOO m4 3 axmtah \ ed aw: . sea;w s .-- ...... ---------- ..... l mmmooz; < :< L 0.. -i 'I' '1- m amunumm :_ cO.EOu+ .33. fibrTmW |"".ll'"IIII'"|'-"--’I'II"II' U sue: a; cauuos mmo: c cmVeOuouuqv ma;an O>uucoo_ asaelm "lll'lII’II"-"IIIIII'III|IIII|.["'I|V. c aw sesame—e a my my o uo_aIOO axmlv mbquOOqz< _ umm + \\\\ l: use: + ca n Pzaumaa ll 3 .m— «aaaaeaaceaceucaeea «« «a.««««««a¢««««hea«aaa«weka«a«a«aaa«tea«tavaaeaacaweaauttacaaea sexes ¢¢««««««« as use II! i i... .II + . "$3332. \ ..z t a + <1...— 3 .1 . . -4tat . «4444 <. «t .- «meet. :::««:««««ztudtt: «teak «eaaaakua _asou I . + Ousnsaz _ Im + m~ use» I muussm a . + «we 3mm + flax lawns + 2 anon masses :usm c . + [Ml newt « 1.! 3...! « :HI; mam: + + 2 I: H; e a x e e a e y IE -ll - Maw- as»; u:— . + :oEm + I #:40uh .33. + « c u:«> un=q . + «We- as“ + a ununv + .~ i¢¥£¢fil¢iiifl {‘ildili£a.“flIi¢¢L§¢ilC‘WK&‘{1*iifl‘ii!!£**£¢i¢i¢fl£¢*{iil‘flifi“! ‘hiiififlfl.«fl««««¢¥¢ ~ch . + due: :fl + a _escu . + Ow=nsaz _ [d + n snow a Ouasoa I sums o . --- -- + «HI 3m.» + ans: aammm + -flm. mm: + + e .33 . .. 2.55.: + -Olm -mmmn + a an q I . -- . Hiatus + . 206 .uuumeu OdOu ass-.5um:=. masons uOquu csuu~cumsa wow smegma; :mx.~m.msw .umsm:< :« :O;o ~__z cans ms asanaauO>nzs use .w:u_~_3 csc. uuluscOOa :91: 1m>ew—E um: uasws—=: u: m~m.wm:s :muq Oxlmslsmamssmc Oshawa sea: annexe pom uq.su ssnmmw .5 Osaka noov Os aelvswz umsxsm was on esteem .O sea: OEOOOA mOAO at. .uosv O:u z: Osaka was: at. ease “aqua ease _.< who: Ono: as —_= .n wads d—< .UUH‘ «~<. Una:— wnfi .3 .awuua mo 1943 _~<. .susz usages» a——s:13um a.s_. assess; Ens: sauna“ Odouuq .mOCa assess: 1a>sm_e ass» sex lax wanna Am>szv .zsmqsuqx cu ass: and» uu>uc an. assocsaumuc Ins: uO>Oc :sw¢:0qz Hugues: ss>sm~snuz luau uvs>w~s cmw.m.a O>M>sz VII. 0 uaazamm Esau: >35 umcu m sea: assume >__wo_wo—ozauoa nuO> .:Ouuumoa Iw :« m« use .1Ommuuazam ma < <1: ”sac. \m>3:\xn OOdflqe 33 :cu s z:_mmwmemsq at tee. .- n. ”has :2 m + Pads 3 asexmum uszxm e.q.m.n.m mxzalm< use n.fi.N.m.m Aumsuv asses; and N.~.~.a.m nz~ mu :wsswo&. mu newsman can so ranch use amused uxsu Oxsu OOnOu III; I!" On [Ill M4=z use; .a30wus see use; an": ass: mccuuuu< anm. awsunov xsua assess—e umxvm Jon use aches sum: autumn—uw< Bonn sacs cssex O ads on onlumvunus us «s.ma .asom use wuss: sag: nasal“: costs: «_<. sum :msslalvusmss anon mes sage: sun: q~< use; szOIumom :samm 0 am ms was mmmzexm 2m- ween uu> was; cut masses :smmm: .>s3e can :smns=. one: smash came: .3 Out uawu :usua .umsu can :«suu 9:9. was; 1:0u smear .s zc_rMm ..wfl at -- amw + m5. mmmm sq :Ossou muss m«\mwsx\ co eeaaaea\eme\ Ban-d a flu mud > H3090 aqqmcouuao usuqsxeoxx O>qu~uun ecu zo—kmuzummn ...: . awe. + z «1.- .1 « I. ens mum: «2 m I< < awzm mac: "cumuumm sacs magalm< . w< umzu nasalm< u=a\w< Mflwm ”4mm "cuwuusm F -< m< d.m.m.m.n n.~.m.m ~.~.~.m.m ~.N.~.®.n ~.~.m.m m._.~.m.m 208 .3o: v—O annex sq Es —. O—Am quw>cs new» as no: _ Ow:c:OuOucu lusmmuz awxewa ~mm dszow mum; see :cummp was 2:12: w>:mum use w ._ ..., zdw .... .zd ...: m mm?“ m... Ha Ha. 3mm -....th . at... . . .ucL—zm c~> ccaumx Os cczsum— Esau Ono: :mmns=. use: ouch mu .O: "n: . cases: o» ccscumn seem usufism an: BOON cases: um um: coguoe cmsuox we :m:eum0 we nmuam Ov—cmu we seem: acumen Na "#1 mazml< w< may}: mewmemwdmwm Ha “3mm an..-» mm. + :3. mm“... mflewmwmm+ flu. ..fimne + mz+elem+ 2.3:: .qooa Osu cu madam =w>evu<. \ Oznmsov Edam doom :« amp >OO>O cs>c=u< :cwaoe cczoxua Mcww uxmuww has as» we: cm>wvum A,m>:al< usa\m< _Ooa ssu cu ceczoup adage .deEO O;b. Osmosa wocacuu doom as usualddsam maquss « :suuce tom rhea excuse hep qwmuvuox Oxlwnwma /c>:a|: use .5::Oum msu =o wc—zd mu saunas Osu uo haon ssh. m. :Oudcu teacuw :o assume misses OumEa:O:~ cu Oo~ use mvmuuo :«Ecn mass wasnw Onccmsn uzmuama O;O mxgalm use _.m.~.m.m m.~.w.n ¢.~.m.n s.m.~.m.m m.n.~.w.m ~.n.~.w.n 209 .cuumu secs 33:: w. _|3ccx : oc<|>=se museums _ amdaeo; Ee:mcus mulacmxmon Oucsamvc; :OE uoaxmnm Owes: uoaxm msmz scum a: o :2 m ":uOuuca sumac “sax Oman—u O>.u.m:sub .mmuucum wcmu~cu one: masseuse: ozy. .c.O .cumoua masseuse: ~anus3u03 new _auu0umuccc Icwuouwuucu was OO»OOL.O vcctumx—a «_mOc= mslummamx mamas: Msudmvcc: .uOOxO 1:: _mom hoaxmamc Ocusomxm< . .08. non "x .ucoa use a—O: p—Oosm O_aswa sewn ash. .£uv> mzu =u tsuoeuuw hosslov cogs usmum noon use O» Auosuv v—Oozn sun» use pOwddsuuqu ImmmmOa sauces amass WOOOOH on Amxvcsama mxmzrc mu accuses nasa|m< Owuscm\m< swam:¢ww mmmewmuwwm seam wmmm s ease + us ca + ez m a .mcwuoso—mx c. Ocuewuueu masses «asu— ssh. sszulqu usuaO—m sown uOuOEO—ud :Ou ~a|uvcu_ Oceans: —=«mO: nave massed“: sec msnuvmco ..o-N ween—Oct semen. can OO—~OOO asses: :v :Ou xswum Alwu.s; use; qcnmccu can was sup xsmnOuAOIOOEAOOV O_ccu:ccwa mzsau< w< cascasncss m>:ala use —.~.~.m.m -.w.r ~.o.~.m.n —.e.~.x.m o.~.¢.n m.m.~.m.m N.m.~.m.m 210 .umxzvm v:qda :euuuws< ass“ On cmumun mama‘s —. _|O>«u new csOuLO§< aupc._; Busuvql wzm dmxncEm wlcsx musswc—m uczu mtwuuo> Os maszdn d OAIOvcccmx :m Ozlsawm On Om.Ec; ::E t l D ) .OO«O> c.cuauax O Ones; _. .1OHOO: mlumu edusu muwuwo> H Eclcqccm «loans» Owens czumpoe use .mecsw anon pecans; ovum “can. ads museum :coc;lw0= pseuuu ZOOuw gowns‘ some cases; .u: passe“! muse unmade =«>uO~ oaumcm .uuoa :Oduwcm assumed con >2. 50:umwu zoom zoolmzt01=ex ma muss = m._._z ease ..=.;; ha; ecu aeaaaa< .tz. edeaeaa: .=.=a Oceaaa< .ez gushes mucoxcm «:mhuue sanmvm .OmOO; .Oe—uscsn a me: muocmOuOsa a: a: Ozo. assume: autumn so“: mnOmOo; u mumwomcmuoua sohu was «mean mxlwcmx :si stemmmuco we use» .aaeaeaaaa :OOOm u—«Oe we; cozucu m.csaec<. m:3 O__On wast—«On noses :snmv~_s:du A—«z neon O_EOOO Ozh. om hzm accuse one; ce:mx psam: :uu<-1—u:3 >_qe:«u calcue misses muncmzch sumac—OJ ueouc OIOEOA <-e was: In: In; H: cOzxmna< ucexm w< .!f;--t.-i. , ;w M.H:, 4?-.i z:.%a.xumm: hzanm use ~OOU WOO—25:0 t< Owccm w< mzscu um:— coacs m. Owed c>:;1;< ace umcu ~.x.~.c.m m.~.c.m ~—.~.~.w.m ~_.~.N.w.n c..m.~.c.n o.~.~.w.n 212 .u—sa Osa amcuws mapped Ozu uza Eomcvmez. use ~_n3 cu Zow~p as culcmnOtumc Ionmcwce In": IIIOI later H O>fiumuc_ use “a F masaum< IOAOH II [I 55".517." III. .I Msez were <-¢ wee: az o a: .ch sou Ouum w O>ms .. oazu OOOOOO» :o—ucs ~1O>O£ so» use aluuuu _ teased O no amount; autumn macm samba; «leaps: see me ccqui magnum "sac ms=z can; a: 9— Oz o a: .pss; me; sad: :OcEOOm ecu :O as u_: o:. a“; u muzueeoum OO masseuse sud: O; OOOOOEOO new cc! OI—OO sass mucus as O use "z 1< .sm_u sq; seq: :ocha use oxen; .~<. was»: zoo :55 Ozscss um: mxzmls< :us=\w< In 5 ' I) I .:u:c5 m.:cmmc= Beau xOczm O;u pOsssuw saucer. Owens»: :mmmm: unsussa Essa ZOO;;|< dcou\w< .mcceeoo Ouch Esau COO Osu asses: H. OOOOOcve —|u;asoa when usaceelso Esau zcc:aua< ~m:3\w< mama—1.3m 2.3 .....3. may... 5:. .. 3mm ‘ PIOOMx us; an: Esau uses as: senu zooclxccsccmmcz a we "Nx ceases— Nc 5mgsumca Oc—usu Ne Nulnmawx came: Ov—cmu um "um —2>:LI3 P8... 3?: 3...: + 27.x was: 37:. 3...: + ....2: 3+ -..... : -m. . n.-.~.w.m N.~_.N.w.n _.~_.~.m.m ~_.N.w.m -.~.m.n 214 .mouucum mud: zeue__:u e_~0u schawzau cum.amu. m:u>wc _:uuu:coa msdas%uoum unansaul uscmuv o>lmalcmumma unvvqusu yam oouawcux c—auacu m:IMmea mamuaa shamans :Nuuasx .95 ca noxcqw a caustOuucq vacuum at. €q_. vans w;:=cuu:. — Cu xuum :«ma tau has sat Zuu vuma a llllill v! + i ll + .Ilo‘!‘ vvuucuqa uwaxmnm< 03:2: vouuwu.v. awaxmnm< hoaxm : hogxm wouacmxm< dcou m< avaxm o< NmJDx amAOm "Chfluufin— Gun—mm 9:40 w>uuuacauudc 3.;auammme lads unv— youasou no: "N: awaxmla< wacax zc_+:—xumx H5 :u—z vzmxmua us: Em —. Es .11ovzouuo asuxuaa scum Ealusx—ov Ealumx Na Ilu‘l 'lip‘ m;mrms NauEu\umc Buss Nausw Ad — A—vEl-Wllxnm‘fl b<+m «I. l .1|.§f‘lll, OI‘I .w#w.tmc:: -tmu; .55:0 5 as v>=w a:a>»a;. w>mw ascu as: _ cu acc>ham ems acaw 39> ca: on ::=>ucn + In: "x zam_a;mw:mww.m .. 2 am: .. _::\ mo_u=mhumwm .1 %; vk<+m + .v.. Mmom gnae< h~u< m m couuau—t mmumm:n us; “a m>zaua< as; + .za—;Euuaguumoa _:c; mauzcm\w< c~.~.x.n n.m_.~.c.m N.m—.~.m.m —.m—.~.m.m m_.~.m.n ¢.e—.N.a.n 216 .swuu a: acovumau: 0:9 osmuw; ucam damn. . asuuwa flow cau~ uaussnou mluougu uvmm —:mm :mu« mzuasao— onmuuu yuan «cw: .cmu— mo uneeqmmh; osu a. uvzm _:wm. a“ ans :cu— u-naaou maumuzu pacm “can .cm»« w:;=::o% uaauo» gnaw «as: ll." '1 .1. m;mz emu; l 11. I. 1| 'tnll. zc—Pn: 13mm:— L2 Auwmv Aguasoav I; uwmu_u=o=_ kuuuuzov— Auumv Auoalmav um uo_u_u:u%— voguqucuv~ P> (in), dar < (in), and 333 (on) are locative and direction markers. On the other hand, 2:. <§ve€>5 (with) is a comitative or instrumental marker (77). Lazard classifies certain expressions which designate physical or psychic conditions as impersonal. Note the following examples (169): 10. sard-am ast cold I is < 'I'm cold' 11. 32 in film xos-am na-yamad from this movie like-me not-came <§e film ne m'a pas pli>> 'I didn't like this movie' Lazard also notices that natural phenomena are expressed by impersonal expressions. From this statement, it appears that the ambient verbs require a non-referent subject (see 3.5.1.). They are similar to the English ambient case frames with the exception that, in English, a dummy subject is obligatory. Some of Lazard's descriptions of case markers in Persian can be matched with those case markers presented in this dissertation (see chapter 3). According to Lazard, 35 (from) is a locative marker which indicates "the point of departure or the origin" (191). Consider Lazard's example: 12. az xane xarej sod-and from home out became-they ééls sortirent de la maisoé? 'They left home' 239 Here, xane 'home' is the place where the action originates. He also assigns 33 (from) to nominals to indicate the passage of the action. e.g. 13. az xiyaban-e lalezar mirav-im from street-E Lalezar go—we (mans allons par l'avenue laleza€> 'We take Lalezar street' (192) In general, Lazard analyzes a number of verbs, though he does not try to classify them based on their attributed subcategorization features. Cases have not been categorized explicitly, but he seems to have distinguished between path, source, goal, and instrument. 4.3.2. Bateni Bateni, in his Description of the Structure of Persian (1969), analyzes Persian syntax based on Halliday's (1961) Categories of the Theoryiof Grammar. His contribution to Persian grammar is a systematic approach to the analysis of Persian syntax. He incorporates the notion of hierarchy into his description of Persian syntax. What be generally fails to do is to include meaning as a factor in establish- ing syntactic relationsips. Bateni presents the following definition of clause (translations and transcriptions are mine): "Clause is that unit of the Persian structure which consists of one or more 'groups' and is used in the construction of higher and larger 240 constituents. In the hierarchy of units, clause is lower in rank than a sentence and higher in rank than 'group'"(74). According to Bateni, a clause consists of four elements. They are: subject, complement, predication, and adjunct (74). In Bateni's analysis of case, there are no differences between subject and agent, and object and range, respectively. He assigns equal values to them since, as he claims, they occupy the same slot (74 fn.). What fills the predication slot is a structural unit which indicates the function of the predicate. According to Bateni, what remains after subject, 'predication' and complement have been taken care of can be categorized as an adjunct (75). As mentioned above, Bateni does not distinguish between subject, agent, experiencer, etc. Any nominal filling a paradigmatic class is grouped as a member of that class without its function being compatible with the one or ones in the same class. Consider the following examples; page references and Bateni's labels for each constituent are presented in the parentheses: l 2 3 l 2 l4. parvin xoshal ast (subject,complement parvin happy is 3 'Parvin is happy' predication:76) 1 2 3 l 2 15. ranande-i otomobil-i dist (subject, complement, driver-a car-a had 'A (certain) driver had a 3 (certain) car' predication:77) 241 1 g 3 1 2 l6. doktor xane nist (subject, complement, doctor house isn't 3 'The doctor is not at home' predication: 78) According to Bateni, Parvin, the driver, and the doctor are all subjects. However, the subject in (14) is an experien- cer; the driver in (15) is a goal since he is a non- transitional possessor of an object; and the doctor in (16) is a patient since the case-frame includes a state verb, which rejects an agent (see 3.5). Bateni categorizes clauses into a major type and a minor type (80—81). He presents the following case-frame (Bateni's formula) for the minor clauses. The parentheses indicate Optional elements: (subject) (complement) (adjunct) predication [i.e., predicate] The following examples illustrate the occurrences of minor clauses: l7. bord took . 'he took it' 18. bord an}: took there 'he took it there' 19. ketEb-ri bord Enja book took there 'he took the book there' 20. javad ketEb-ra bord Enja Javad book took there'Javad took the book there' His formula, however, should be rearranged as follows: 242 subject, complement, (adjunct), predicate that is, both the complement and the subject, though covert on the surface, are obligatorily present in the underlying case-frame. The agent's presence is implied by the zero morpheme subject marker in (17), (18), and (19). On the other hand, the predicate bordan (to take; to carry) necessarily requires a patient, which might be covert on the surface but understood in context. In general, Bateni's analysis does not present a deep-surface relationship of the constituents on the syntag— matic axis. 4.3.3. Khanlari One of the most comprehensive and systematic studies of the evolution of Persian from Old Persian to modern Persian is by Khanlari(l975).10 He makes a diachronic study of different aspects of the Persian language. Never- theless, like many of his predecessors, most of his analyses are exclusively concerned with morphological and syntactic features. The major sources used are classical and literary works, though he points to certain colloquial uses of the language from time to time.11 Khanlari classifies verbs into the three major sub- categories of gozara (transitive), lazem (intransitive), and nagozar (non-passing) (vol. II:320). In the third category, verbs or actions affect only the subject. In 243 other words, though these verbs, like transitive verbs, require an object, the action has no effect on the nominal objective which preceeds the verb (vol. II:320). A significant morphological characteristic of the so-called non-passing verbs is that the formal subject does not agree with the subject marker suffixed to the verb. Thus, the verb almost always agrees with a third person singular (320). In a sentence such as (vol. II:325): 21. xos-Em Emad happy—me came-it (lit. it came well to me) 'I liked it' Khanlari considers the surface :am (me) as the logical subject (nehad)of the clause (Windfuhr 1979:126). Many of ambient experiential case frames presented in this disser— tation (see 3.5.2) can be applied to this category, or vice-verse. Thus, Khanlari's examples: 22. sard-am ast 'I'm cold' 23. garm-am ast 'I'm hot' can be assigned the following case-frame: S—Ambient Experiential in which :gm (me/I) is the experiencer. There is no agent/ subject to agree with the predicate (it is always in third person singular form). This case-frame is not restricted 244 to the state verbs lexicalized with the verb budan (to be), but can be expanded to include process verbs as well. Consider the following examples: 24. dire sard-am mise 'I am getting cold' 25. dare garm-am mise 'I am getting hot' (see Chapter 3 for more details). In general, role-predicate relationships have not been fully dealt with in Khanlari (1975: volumes I, II). 4.3.4. Lambton Lambton (1976), like many of her predecessors, makes a general analysis of Persian. Her observations are valuable, though they are not insightful enough to allow the development of a semantico-syntactic analysis. Her analysis of cases is restricted to surface realizations of instrument and place in Arabic (196-199). As far as Persian is concerned, she only mentions a direct object case which. is superficially marked by the postposition 25;. This indicates that Lambton, as well as many other Persian grammarians, has not paid enough attention to case-predicate relationships when cases do not have representations on the surface. 245 4.4. Transformational Grammar and Persian 4.4.1. Palmer (1971) Palmer is the first scholar who applied Fillmore's 'case-grammar' (1966) to deal with 'Ezafe-construction' in Persian. He applies rules to relate the underlying cases of Persian with the linear linguistic constituents on the surface (Windfuhr 1979:43). He suggests modifica- tions of Fillmore's case theory in order to adapt it to Persian syntax. Palmer sets up five cases: agentive, dative, instrumental, objective, and locative (32). Palmer's cases partially overlap. Thus, according to him, what makes a dative case different from an objective case is the fact that in the latter, the animateness is not specified. However, in both cases, the nominal associated with a predicate is affected by an action or a state (32- 33). Palmer's locative case can be divided into locative and goal as presented in this study (see Chapter 3). Palmer's objective case, like that of Fillmore's, is a waste—basket which includes anything excluded from the remaining cases. Since Palmer's deep structure encompasses a set of phrase structure rules, the underlying constituents are basically linear in order. These rules, when applied to the underlying semantic nodes, reinterpret or rewrite them into subconstituents in the order defined (34-35). 246 No priority is given either to the nominal or to the predicate within the case frame in determining the other. 4.4.2. Marashi Marashi (1970), following Chomsky's Aspects model (1965), demonstrates the interrelationship of verbs and other syntactic structures. He assigns two major con- stituents to a sentence: a Noun Phrase and a Verb Phrase, the verb being the core of the latter (3:32). Marashi shows that a morphological analysis of surface structure by itself does not provide an adequate basis for the analysis of verbal constructions. He cites examples such as: 26. man xeyli dust dar-am I much/many friend have-I which can be interpreted either: a. I have many friends or b. I like (it)very much Marashi sees the difference, and states that if interpreta- tion (a) is understood, the verb has two constituents, whereas in (b), the nominal dust and the verb dastan are inseparable and should be considered a single constituent (6). This ambiguity could be better explained by postula- ting different underlying cases for the surface subjects. Thus, in (a) man is a goal and dust is a patient, whereas in (b), man is an experiencer and the partially covert :2 (it) is a range (see Chapter 2 and 3). Marashi's observation 247 is another indication that underlying rules should be sought rather than surface rules. Marashi classifies verbs according to whether they are a) transitive or intransitive, b) stative or non-stative, and c) motion or non-motion (13). Thus, gozastan (put something somewhere) is a transitive, non-stative, motion verb; raftan (go) is an intransitive, non-stative, motion verb; and xandidan (laugh) is an intransitive, non-stative verb (13). According to Marashi, what differ- entiates a stative from a non-stative verb is that the former does not undergo the progressive transformation (similar to Chafe's 1970 test of stative versus non—stative verbs) (23-24). Marashi assigns lexical features to verbs in their underlying representations. Verbs are semantically loaded with features which indicate the how's and when's of the application of transformational rules. They also specify the types of subjects which can semantically collocate with verbs (23-24). 4.4.3. Bashiri Bashiri (1972) develops his own 'deep' system of case-frames. He attempts to substantiate his theory through reference to Persian. Bashiri, in discussing the relations between notions such as agent, source, path, etc. 248 with the predicates, develops a set of complicated rules. Because of the originality of his theory, a cursory discussion might not do justice to Bashiri's analysis. The reader should refer to his work for a more comprehen- sive treatment.12 According to Bashiri, existence is the most essential element of a sentence; that is, budan (to be), in its ontological sense, is the core element occurring between the Real World and the linguistic primitives in a hier- archical manner(49). The properties of existence are pro- jected into the surface by the attribution of BUDAN (to be) to either nouns or verbs. Certain rules (mapping rules and correlative rules) apply to derive superficial constituents of clauses from their deep representations. Functions, in Bashiri's analysis, are created by mapping predicates to nominals (110-111). The reapplication of these rules in a consistent manner leads to new nominals and new predicates. In Bashiri's analysis, there are four types of sentences: active, passive, factive, and causative (165- 66). The factive sentences are characterized by the verb budan (to be). Bashiri sets up six cases, which are ordered in a hierarchy with respect to their importance in undergoing l . . rules: agent, source, path, 3 experiencer, goal, and object. 249 If my interpretation of Bashiri's analysis is correct, then it appears that Bashiri, in assigning case labels to the nominals, has been too much influenced by the surface markers. Although he emphasizes that the path-marker is not always bi (with) in Persian, his examples tend to illustrate the opposite. Consider the following examples (58): 27. husang bi man sohbat kard Hushang with me speech did-he 'Hushang talked to me' 28. husang b3 kard §1r kost Hushang with knife lion killed-he 'Hushang killed the lion with a knife' According to Bashiri, both the underlined nominals in (27) and (28) are paths. However, with the definition he provides for path, the nominal in the first sentence can hardly ever be a path. mag in (27) is an experiencer, whereas, EEEQ (knife) in (28) is an instrument. Other examples are as follows (55): (The underlined nominals are sources in Bashiri's terminology). 29. dolat be u hoquq midah-ad government to him salary gives-it 'The government pays him his salary' 30. soqrEt az zahr mord Socrates from poison died-he 'Socrates was poisoned' 250 31. ali az hasan pul gereft Ali from Hassan money took-he 'Ali borrowed money from Hassan' 32. husang az beyn raft Hushang from among went-he 'Hushang died' A finely grained analysis which takes into account more factors in the analysis of cases indicates that dolat 'government' in (29) is a potent agent/source; zahr in (30) is an instrument; Hassan in (31) is a source from whom something is taken away (Sullivan's 'maleficiary' (1980)) and az beyn raft in (32) is a compound verb whose elements are inseparable (see Khanlari 1975 vol. II:3l4-319).14 See Chapters 2 and 3 of the present work for more details. 4.5 Conclusion Case relationships in Persian have so far been either ignored or only tentatively analyzed. Syntax has been the focus of attention, with few efforts to relate it to semantics as the level of language closest to the world of reality. In addition, most examples have been taken from classical Persian poetry rather than the colloquial language. Jones, Forbes, Phillott, and other traditional grammar- ians make efforts to produce detailed descriptions of the Persian language in general. What they generally have in common is that in their descriptions of language, they 251 mostly ignore meaning. Phillott is, of course, an exception since, in his defined classifications of verbs and cases, he partially incorporates meaning into his description. However, none of them make a realtively dichotomous dia- tinction between underlying representations of cases and their superficial representations. What we presented in Chapter 3 of this dissertation is different in the sense that I attempted to provide more factors in the realizations of cases. Moreover, based on Longacre's model (see Chapter 2 of the present work), I established more direct relationships between surface and deep structure constituents. Notes to Chapter Four 1The term 'Iranist' refers here to the European scholars who have worked on Persian grammar. 2The term 'traditional' is attributed to those linguists or grammarians (pre-transformational or post-transformation- al) who deliberately did not base their analyses on the transformational theory. For a more descriptive analyses of the Iranists' and the Iranian grammarians' contributions to the Persian grammar, see Windfuhr's comprehensive Persian Grammar (1979). The use of :EE is very controversial. Some scholars consider it as a direct object marker, whereas others consider it both as direct and indirect object marker (see Windfuhr 1979:41-57). According to Windfuhr, "...the impression created by frequency that £3 is only an object marker is misleading" (49). He also states, "Peterson argues first that £3 is a marker of topicalization (i.e. theme) and second that the necessary and sufficient condition for its use is not that the noun phrase be definite but rather that it be specific, i.e., that it have, from the speaker's point of view, a unique referent"(56). 5Poetry has been and still is the major source of data collection. 6According to many grammarians, passive is expressed by the participle and the full paradigm of £31 'become' (Windfuhr 1979: 105). For a transformational explanation of passivization in Persian, see Moyne 1974. 7Here are the descriptions Phillott presents for the case markers in Persian: Nominative: " ...it is often employed at or near the beginning of a sentence in an absolute sense to introduce the subject, being independent of the grammatical construction that follows it" (445-6). 252 253 Genitive: It is marked by ezafe, e.g.: pesar-e malek 'the king's son' son-E king Dative: He presents different markers, including bi, as in ba u_goft-am 'I told him'. He also mentions that the use of E; adjacent to the dative case is obligatory when it specifies possession (448). His examples are as follows: t3 harke tir az halqe-ye angostar-i bogzaranad xatam u-rE basad mara baq-i-st me garden-a-is 'I have a garden' Accusative: it is marked by -ra. e.g.: ketEb-ra be man bede book-Ach to me give 'Give me the book' Vocative: -y3/ey Ablative: az -- 8Corroborative apposition "in word" refers to the repetition of the formal head noun in the clause. The corroborative apposition "in sense" refers to the case where a word emphasizes the selfhood or totality of the head noun. The verb raftan 'to go' is ambiguous, since it does not make it clear whether the agent initiates and performs the action or the action is important and not the performer. In other words, u miravad 'he goes' does not show whether 3 'he' 'walks' or 'is taken by something'. 10The first edition appeared in 1959 and has been reprinted six times. 11Apparently, there is or should be a third volume which develops more studies of different aspects of modern Per- sian. I have seen no reference to it in any text available to me. The author's preface to the second volume anticipates another work by the author, however (l975:8). 254 12 To demonstrate how a sentence containing various notional categories is generated, the following sentence will be analyzed based on Bashiri's system of analysis: 1. b3 from hatred & anger with finger-pl-E V piremard az nefrat o xasm panje-ha-ye old man xod gelu-ye mard-e servatmand-ra fesord self throat-E man-E rich pressed 'The old man, with hatred and anger, pressed the rich man's throat with his fingers/hands' NBU VBU r “““‘~\“‘ BU v"/ ‘\\\\““BU //’:j \ \\\ / \\ ’//' ¢ ’,,z// \¢ / //” '1 N \ / \ AG ,_ - ALV P \ ' ° SOURCE SLV PATH ELV . \//1 ' EXP ELV I I % AG S P EXP ELV AG = piremard BU = Budan 'to be' S = az nefrat o xasm ALV = Agent Level Verb P = b5 panjehiye xod PLV = Path Level Verb EXP - geluye marde servatmandra ELV = Experiencer Level Verb ELV a fesord SLV = Source Level Verb As the diagram shows, in Bashiri's system, the logical sum of g and 2 creates the Agent. mapped onto V, When Agent is correlatively an agent level verb is produced. The other rules when applied in order are as follow: 255 1. Agent level Kerb U Agent -——-———) Source (to be read: agent level verb mapped onto agent creates source) 2. source /9 ALV ; SLV (source level verb) (to be read: source correlatively mapped onto ALV creates SLV) 3. source L} SLV -——————9 Path 4. Path (A) SLV ———————9 PLV (path level verb) 5. Path k} PLV —————-—9 Experiencer 6. Exp {1 PLV ——-—-——9 ELV (experiencer level verb) 1323£h in Bashiri's system is vague. Bashiri's definition for path is as follows: "Path is created by mapping the SLV [i.e., Source Level Verb] into the source" What Bashiri apparently calls path is experiencer and instrument in Longacre's system, commitative in Fillmore's (1968). In the present work path refers to: a. the locale(s) transfersed in motion predication; and, b. transitory owner Consider the following examples: a. John travelled from X to Y via Z b. John gave a book to Mary for you. 14The sentence: 2. husang az beyn raft Hushang from among went is an ambiguous sentence and can be interpreted as in (a), and (b) below: 2a. Hushang died 2b. Hushang left us (lit. he went out from among us) If (b) is understood, then 32 beyn can be interpreted as as source. ‘ Chapter Five Comparison of Persian and English 5.0. Introduction In the preceding chapters, we examined some of the features attributed to the case frames of Persian. Following Longacre's model, we categorized verb forms in Persian according to different parameters based on the quality and number of features each individual verb possesses. We examined the relations among cases (roles) and their related verbs at the predication level and cited examples in which the number of constituents in the semantic con- figuration was different from the number of constituents at the surface structure level. This implies that utter- ances have an underlying organization which is substantially different from their surface structure. In this chapter, we intend to illustrate a comparison of case frames in English and Persian and to make a brief analysis of the semantic and syntactic differences between the two systems. All hypotheses presented in this chapter concerning the emergence of problems are tentative,and further research would be necessary to substantiate them. To develop a framework against which this analysis can be made, we made the following assumptions: a. Longacre's analysis of English case frames truly represents the deep structure of the language; 256 257 b. the underlying structure of the language is the most basic linguistic structure, corresponding closely to the organization of thoughts in the mind of the speakers of the language; c. the basic case relationships are universally applicable; d. following Longacre, the surface structure, although primarily the domain of form, has meaning correla- tions, for example focus and topicalization (301); and e. the deep structure consists of a number of un- ordered constituents, the surface order is langu- age specific, and its transparency versus Opaque- ness; elasticity versus non—elasticity; and rigidness versus flexibility varies from language to language. In other words, English and Persian allow different degrees of flexibility and rigidity in their surfaceability. In comparing the two systems of case frames in English and Persian, and also the deep structures versus their related surface structures, we arrive at the three possible relationships diagrammed here: 258 LEnglish URI Persian UR [_English Lfersian IEnglishJ [Persianj [Englishj [Persianl Eng. surface English Persian English [Persian Pers.surface surface surface surface surface A B C A represents cases in which one identical or similar concept surfaces as syntactically similar strings of constituents on the surface of the two languages. In such cases, the surface word order in the two languages would show a similarity in linearity, which is a feature of surface structure but not deep structure. As far as my data show, the English and Persian surface instransitive clauses fall into this realizational category. Thus, we have the following examples (arrows indicate deep-surface relations): 1. DIE {P—Phys P —-——-) Ali died 2. SINK motion {P-Phys P -—————-> The ship sank 3. GET SCARED P-Exper E (I:«—-——9 Ali got scared (inst) 259 la. MORDAN {P-Phys P } ali mord 'Ali died' ali died 2a. QARQ SODAN kesti qarq P-Phys éod 'The Ship P fi. ship sank sank' motion 3a. TARSIDAN ali tarsid 'Ali got P-Phys I A v (inst) E ( ) —7 Ali feared scared Examples in this category are restricted to case frames in which a predicate is accompanied by at most one argument. The major problem occurs when similar predicates surface as simple verbs in English but as so-called 'compound verbs' (see 3.1.7.1) in Persian. In examples (1), (2) and (3), the basic case frames are translatable into their primary non-configurational un-marked forms in both languages. The elaboration of the basic entities may lead to surface forms which are semantically marked. Here, we should take into consideration the richness of one language versus another in the number of morphological variations in the surface structure. For instance, Persian allows a choice of stylistic alternations related to example (la) above. Thus, we have the following constructions semanti- cally marked for (la): lb. ali rehlat kard (if Ali was a saint) Ali death did 260 1c. ali sahid sod (if Ali was a martyr in Ali martyr became a holy war) 1d. ali dar gozast (an official announcement) 1e. ali gurtizid (if Ali was a 'bitch') Ali died Although these examples differ on the surface as to emphasis, style, etc., they all have the same case relationships. However, in producing literary translations, these sarcastic and stylistic uses should be taken into account. p represents cases where two similar concepts surface differently in the two languages, not only in terms of their linearity of constituents but also in the number of surface clauses. Take the following as an example: 4. GIVE AP-Phys motion A/S Path P G -————§ Ali gave Reza a book n possessio for Mandana I *Ali Reza gave a book for Mandana 4a._DADAN AP—Phys motion A/S Path PG -———) a) Ali yek ketEb be possession reza dad ke be-de be mandana S—give to mandana Ali gave Reza a book for Mandana b) ali be reza vek ketab dad ke be-de be Mandana Ali to Reza one book gave that S-give to Mandana Ali gave Reza a book for Mandana 261 c)? ali yek ketib be reza dEd Ali one book to Reza gave baraye mandana for Mandana Ali gave Reza a book for Mandana Here, Persian has the tendency to represent case frames of this type by a matrix clause with an embedded clause in juxtaposition to it. In (4a), two clauses represent a single proposition. Examples of this type are common in other languages in which, for instance, a clause such as the English You told me surfaces as two related clauses: You said I heard (Longacre 1976: 302). 9 represents cases in which concepts in the two languages have both different semantic prOperties and different surface repre- sentations. Predicates of this type are mostly idiomatic, frozen expressions. No particular case frames can be devised for them, and their analyses are subject to further elaborations. The underlying concepts are different and no one-to—one relationship can be formulated for the underlying constituents with those representing them on the surface. Some of these concepts are either totally absent from English or have only vague analogs. These examples were excluded from the data. However, to clarify this point, we present the following: 262 5. m3 nokar-e somE—im 'I'm at your service' we servant-E you-we are (lit. we (i.e.,I) are your servant) 6. az dast-e yek xinom-e xoskel zahr-r3 ham na-bEyad rad kard from hand-E one lady-E beautiful poison-Ach also neg- should refuse 'One shouldn't reject even poison from a beautiful woman's hands' Neither (5) nor (6) can have case frames in which the arguments fit their predicates. The meaning conglomerate of the surface grammatical forms does not represent identi- cal semantic concepts. Both (5) and (6) are complements1 and the concepts related to them are interpreted in several different ways when translations are presented to native speakers of English. What a native speaker of Persian understands from examples (5) and (6) is probably different from the impressions a native speaker of English gets from the translations.2 Fifty-five examples of this kind were excluded from the analysis. 5.1. English Case Frames and Persian Case Frames: the emergence of differences In this section, we emphasize the nature of differences in case frames when the two systems are compared. Neverthe- less, the congruity between the two systems of Persian and English case frames indicates the fact that the two languages involved are closer in their deep structure configurations than their surface representations. We only concentrate 263 on parameters and case frames which differ sharply in their semantic configurations. Following a discussion of differ- ences in the case frames, examples are provided of differences in word order, ellipsis, cognate prolongation, and clause duality. 5.1.1. Ambient Case Frames As already shown (see 3.5.1), in English state, process, and action—process ambient case frames, the occur- rence of the empty ip is more common than the occurrence of the general nominal weather. In other words, English shows a tendency to substitute 13 in subject positions of ambient verbs in general. The action ambient case frame is different in the sense that the nominal weather apparently cannot fill the subject slot. Thus, paraphrases of examples (7a) and(8a) are (7b) and (8b), respectively. However, (9b) as a variant of (9a) is unacceptable. 7a. It's hot today 7b. The weather is hot today 8a. It's warming up today 8b. The weather is warming up today 9a. It's raining/snowing 9b. *The weather is raining/ snowing In Persian, the nominal hay; 'weather' is most likely to occur in the subject slot in such cases. However, its absence in everyday speech is more common. When this occurs, no substitute is necessary to fill the gap. Examples of action-process case frames are rare. 264 5.1.2. Ambient Experiential Case Frames Examples of state ambient experiential case frames presented by Longacre (1976:44) are ambiguous. That is, they fall into different groups based on the interpretations obtained. Consider the following examples: 10. The patient is hot 11. The patient is cold Example (10) can have the following interpretations: a. The patient's body temperature is above normal b. The patient is hot probably because the weather is hot c. The patient's body is hot as it is experienced by someone else If c is understood, then the case frame is state physical and the patient is a patient rather than an experiencer. lo and 2 fall in the experiential case frames. In Persian, the predicates morphologically differ with each of the interpretations above. The following correspond to 10a, 10b, and 10c, respectively. The related case frames follow each example. 11a. mariz tab dar-ad {S-Exper} patient fever has-he E llb. mariz garm-es-e S-Amb patient hot-him-is Exper E llc. badan-e mariz daq-e {S-Phys } body-E patient hot-is P 265 5.1.3. Experiential (Completable) Case Frames The English case frames in these groups are comparable to those in Persian. However, the eixstence of homo- phonous verb forms in Persian, and for English in certain cases, make translating from one language into the other problematic, at least to new learners of the language(s). Note the following examples. Case frames refer to experiential (completable) case frames in both languages: 12. Tom got scared P-Exper (inst) E(I) 13. Tom scared me with a AP- Exper firecracker intention A E (I) (inst) 14. Susan learned English P- Exper complet E R 15. Susan taught her students AP- ~Exper English complet A E R 12a. ali tarsid -Exper Ali scared inst) E (I) _ _ 4' 13a. ali mara tarsand AP- -Exper intention A E (I) (inst) J 14a. maryam ingilisi amuxt P- -Exper Maryam English learned complet E R I 15a. ali be sigerd-Zn-as ingilisi amuxt [AP- -Exper Ali to student-pI-his 1 complet A E R English taught ' In (15), the occurrence of the experiencer is optional. In (15a), the experiencer cannot be covert and must obli- gatorily surface. This is different from (14a), in which 266 the experiencer fills the subject slot. In other words, in action-process experiential completable case frames (15), English but not Persian allows the suppression of the experiencer. However, Persian verb forms are morpholo- gically different if the experiencer is Optionally suppressed. Thus (16a, b, and c) are the corresponding forms of (15) if the experiencer is suppressed. 16a. ali (be sigerdanas) ingilisi dars dad Ali (to his students) English lesson gave-he 16b. ali (be sagerdanas) ingilisi ta'lim dad Ali (to his students) English teaching gave-he 16C. ali (be sagerdanas) ingilisi yad dad Ali (to his students) English taught 5.1.4. Experiential Directed Completable Case Frames In English, the occurrence of complement (range) in sensation and speech predications is optional. In Persian, however, it is necessary to specify nouns such as sound and pgp£_in process and action-process predications pro- vided that they are not lexicalized in the verb forms. Thus, one does not hear a bird, but its voice, and does not smell garlic but rather its odor. Take the following examples into consideration: 17. ali bu-ye sir sanid {P-ExPer}ERS 'Ali smelled ' Ali odor-E garlic heard complet garlic 18' bu-ye sir mi-ya-yad 'A-ExPer}ERS 'One can smell - v odor—E garli pres-come-itlcomplet garlic 267 19. gol—ra buid o goft bah A-Exper A/E S 'Smelling flower-Ach smelled and complet hzesiizwer said nice _ 'It's nice" 20. *bu-ye gol-ra buid o goft bah, smell-E flower-Ach smelled and said nice In example (19), the complement (range) is incorporated into the verb form. The optionality of range in English case frames of this group versus its obligatory presence in the Persian counterparts does not seem to pose any major problems for learners since English allows both forms. 5.1.5. Physical Motion Case Frames The collocation of source, path, and goal in Persian produces examples which are 'heavy' to native speakers' ears. An equivalent of example (21) would be the less common examples (22a,b,c) and the more common example (23): 21. Tom carried the basket from the kitchen through the dining room into the living room (Longacre 1976:49) 223. (?) ali sabad-ra az Espazxine az tariq—e otEq-e Ali basket-Ach from kitchen from way-E room-E nEhErxori be otEq-e nesiman bord dining room to room-E sitting took 22b. (?) ali sabad-ri az tariq-e otEq-e nEhErxori az Ali basket-Ach from way-E room-E dining room from Espazxane be otEq-e nesiman bord kitchen to room-E sitting took 22c. (?) ali sabad-rE az otEq-e nihirxori be otaq-e Ali basket-Acc from room—E dinning to room-E 268 V —v - nesiman az tariq-e aspazxane bord sitting from way-E kitchen took 23. ali sabad-rE az aspazxane bar dist va az tariqe Ali basket-Ach from kitchen took and from way of otEq-e nEthxori be oth-e nesiman bord room-E dinning to room-E sitting took (carried) My informants unanimously accepted (23) and gave some preference to (22b) with respect to (223) and (22c). 5.2. Word Order What English allows to be done through different arrangments of clause constituents, Persian conveys through. case markings. In other words, since ranges, patients, goals, etc. are superficially marked as objects and complements, permutations are less likely to change meaning as they do in English. Compare the English clauses (24) and (25), in which the different order of constituents changes meaning, with the Persian examples (26), (27), and (28). 24. The cobra swallowed the porcupine 25. The porcupine swallowed the cobra 26. mar-e kobrE jujetiqi-ri bal'id snake-E cobra porcupine—Ach swallowed 27. jujetiqi-ri mar-e kobra bal'id porcupine-Ach snake-E cobra swallowed 28. mEr-e kobrE bal'id jujetiqi—ri snake-E cobra swallowed porcupine-Ach 269 (26), (27), and (28) have meanings equivalent to that in (24) but not (25). (28) is odd, though acceptable. These different arrangements of constituents, although not substantially changing meaning, divert focus from one element to another in the construction. In (26) (unmarked order), mare kobra 'the cobra' fills the subject slot and the rest of the clause is the new information for the tOpic given. (27) is patient-focused, whereas (28) is action- focused. 5.3. Superficial Morphological Differences Another problem which at least shows a superficial divergence between English and Persian is the morphological realization of the same predicate by two different lexical entries. The predicate features determine the verb forms and also their relationship to the arguments. Where English shows a tendency to use identical verb forms, Persian pre- sents different verb forms in similar case frames. This difference in surfaceability probably gives rise to conflicts and mis-interpretations in language teaching and learning. The following examples illustrate this point: 29. cub-r3 zad-am be narde stick hit-I to fence 'I hit the stick against the fence' 30. cub xord be narde stick collided to fence 'The stick hit the fence' 270 31. *narde-rE b3 cub zad-am fence with stick hit-I 'I hit the fence with the stick' 32. *cub zad be narde stick hit to fence 'The stick hit the fence' 33. ali tuye xiyEbun be man tane zad Ali in street to be bumped 'Ali bumped into me in the street' 34. ali tuye xiEbun xord be man Ali in street collided to me 'Ali bumped into me in the street' Example (29) is different from (30) in the sense that in (29), the agent is Optionally omitted on the surface. (30) has no agent in its case frame. Moreover, xordan (lit. 'eat' in its configurational sense) never collocates with an agent. Since no sentient agent exists in the case frame of example (30), the predicate xordan has no feature/ intention/ in its semantic configuration, which differen- tiates it from zadan in example (29). However, as the English translations show, English tends to represent both by one verb form. (31) would be acceptable if an animate nominal were substituted for narde 'fence'. (32) is unacceptable since Egg 'stick' can neither be an agent nor a potent agent. The factors which differentiate between (29) and (30) are also relevant to (33) and (34). English has some special surface structures in which an intransitive verb form occasionally functions as a 271 transitive verb. On the contrary, in similar cases Per- sian morphologically distinguishes between the two. Consider the following examples: 35. I walk in the park every evening. 36. I'll walk the dog for you and compare them with their Persian equivalents: 37. man har ruz dar park rah mi-rav-am I every day in park walk 'I walk in the park every day' 38. man sag-etan—ra baray-etan mi-gardEn-am I dog-your for-you walk 'I'1l walk your dog for you' In certain grab, acquisition, and transfer case frames, English structure allows the suppression Of one or two arguments without requiring morphologically different verb forms. In other words, the cases (roles) may Optionally be covert on the surface structure with no difference in the accompanying verb forms. Persian does not allow the omission Of arguments in similar constructions unless verb forms with incorporated nominals are used instead. Here is a comparison of some related constructions in English and Persian: 39. John gives money to the poor. 40. John gives very generously. While (41) is not acceptable in English, (42) is apparently good English: 272 41. *John gives tO Susan 42. John gives to the poor inhabitants of ghettos. The fillers Of goals here are restricted to nominals such as: the poor, the needy, etc. (Platt 1971:84). However, the verbs are morphologically distinguished in Persian. 43. ali be foqara pul mi-dah-ad Ali to the poor money pres-give—he 'Ali gives money to the poor' 44. ali sexavatmandane mi-baxs-ad Ali generously pres-give-he 'Ali gives away (his property) generously' 45. *Maryam be ali mi-dah—ad (bad connotation) Maryam to Ali pres-give-he 46. *xEnom-e ahmadi be sEkenin-e que-ha mi-dah-ad Mrs. Ahmadi to give-she 'Mrs. Ahmadi gives 47. xanom-e ahmadi be kon-ad Mrs. Ahmadi to gives (bad connotation) inhabitants—E ghetto-pl pres- to the poor inhabitants of ghettos' sakenin-e zaqe-ha enqu mi— inhabitants-E ghetto-pl 'Mrs. Ahmadi gives to the poor inhabitants of ghettos' 5.4. Cognate Prolongation Certain English predicates are accompanied with nominals which function as nominal cognate prolongations of the predicates themselves (Longacre 1976:29). The purpose is probably to specify more completely the activities which have already been specified by the predicates (Longacre 273 1976:68). Nominals of this kind are normally related to the same roots as their predicate forms. These nominals can further be elaborated and expanded. Note the following examples: 48. The Beetles 232g beautiful songs of peace 49. The Iranians fought (a good fight) against the Iraqi troops Persian does not allow prolongation of predications either of the cognate prolongation type or of the generic nouns semantically associated with the predicates themselves. It is interesting, however, that Persian has simple and compound verb forms to realize the same concepts in cases where English allows the elaboration Of cognate prolonga- tions. Consider the following examples: 50. marziye dar jasn-e arusi-ye baradar-am avaz xand Marziye in party-E wedding-E brother-my sang 'Marziye sang at my brother's wedding party' 51. irani-ha b5 etzqi-hz jangid-and Iranian-p1 with Iraqi-pl fought 'The Iranian troops fought against the Iraqi trOOps' 52. irEni—ha b5 eriqi-hi japgre mardane-i kard-and Iranians with Iraqis fight-E courageously-a did- they 'The Iranian trOOps fought a courageous fight against the Iraqis' 53. *irEni—ha b5 eraqi-hi fang—e xubi japgid-and Iranian-pl with Iraqi-pl fight-E good fought- they 54. ali xandid 'Ali laughed' Ali laughed 55. 311 xande-ye boland-i kard 'Ali laughed a Ali laugh-E loud-a did hearty laugh' 274 56. *ali xande-ye boland—i xandid Ali laughter-E loud-a laughed 5.5. The Conflict in Lexicalized or Incorporated Verbs We have surface structures in English in which an instrument is lexicalized or incorporated into the verb. In other words, the incorporation of instrument into the predicate makes the repeated occurrence of the instrument redundant. Consider the following examples in English: 57. John knifed his wife. 58. *John knifed his wife with a knife. and now compare them with the following Persian examples: 59. hasan zan-as-ra cagu zad Hassan wife-his knife hit 'Hassan knifed his wife' 60. hasan zan-as-ra b3 EEqu zad Hassan wife-his with knife hit 'Hassan knifed his wife' 61. *hasan zan-as-ra caguid 'Hassan knifed his wife' As seen, a simple instrumental incorporated verb form is not allowed in Persian. In compound verb forms, the instrumental marker can Optionally occur in juxtaposition to its nominal. However, there are examples in which the optional selection of the marker leads to ambiguities. Consider the following: 62. hasan sar-e ali-ra tig zad Hassan head-E Ali-Ach razor hit 275 63. hasan sar-e ali-ra b5 tiqézad with (62) has only reading (3) below, whereas (63) is an ambiguous clause and can mean either (a) or (b): a. Hassan shaved Ali's head b. Hassan cut Ali's head with a blade 5.6 Integrated Clauses Here we refer to the conflation of two or more predi- cations so that they surface as a simple sentence. In constructions of this type Persian allows the ellipsis of identical nominals by incorporating the particle kg 'that' into the construction. What differentiates English from Persian with this respect is the number of possible inter- pretations in Persian. Consider the following examples with their related underlying case frames: 64. I shouted to Ali for Hassan to leave. with only one interpretation that 'I shouted to Ali so that Ali tells Hassan to leave'. The expanded case frames are as follows: {fig-Exper} AP-Exper Ai Ej + A-)Aj rected A Ei + directed r complet Now compare (64) with its equivalent in Persian (69) which conveys two interpretations as in (653) and (65b): 65. sar-e ali did kesid-am ke hasan be-rav-ad head-E Ali shouted-I that Hassan S-go-he 65. a.{AP-Exper] A E + AP-Exper A E + A— A 'ldirectedj i directed i j j L complet ) f 1 ' ) 65b. JAP-Exper A- I Idirected A E1 Ej + { jA/Ej 'I shouted to Ali so that Hassan hears me and leaves' Examples similar to (69) are always ambiguous in Persian and may lead to misunderstandings. 5.7. Conclusion In this chapter, we have discussed some areas of controversy which arise when the two systems of case frames for English and Persian are compared. We presented examples in which similar semantic configurations lead to differences in surface structure because of the divergent manner each language relates semantically-tied constituents to their correspondent surface realizations. Some of the examples we provided showed where the two systems of case frames are similar or dissimilar in their surface structures. The differences noted are simply the by-products of any two case frame systems when they are compared. Whether the tentative hypotheses presented here concerning the language teaching and translation can be verified or not must be the subject of further systematic, experimental, and analytic research. Whether some divergencies are more crucial to language teaching is beyond the scope of the present work. Some of the differences discussed here are unilaterally significant in language teaching. In other words, some problems are relevant to learners of Persian, others to learners Of English. For instance, the 'dummy' t in the 277 English ambient case frames is probably more unusual to Iranian learners of English than the use of have 'weather' as the filler of subject slots in the Persian ambient case frames is for English-speaking learners of Persian. Notes to Chapter Five Compliments are expressions and turns of phrase which are commonly used in everyday conversations. They are either semantically void or formalities. 2My judgment is based on the impressions my English informants got from the translations. 278 Chapter Six Summary, Limitation, Extension 6.1. Introduction This study has presented a descriptive analysis of the Persian case frame system. The main purpose has been to classify the Persian case frames and assign roles to participants within the predication. An effort has been made to establish the closest possible semantico-syntactic relationships between predications and their superficial clause representations in Persian. For this purpose, Longacre's (1976) model of deep—surface classification of language was adopted as a theoretical model. Longacre's analysis was used as a qualitative reference, and an analysis of Persian similar to his analysis of English was conducted. Furthermore, his analysis of English case frames was partially matched with those case frames of Persian developed in the present study in order to discover the most significant differences between English and Persian at the predication level. In order to collect the data, the novel Daijan Napeleon by Iraj Pezehkzad was used as a quantitative reference. Indicative clauses were selected, and features which were attributed to each individual verb were speci- fied. Roles were assigned to nominals accompanying verbs, 279 280 and each clause was given appropriate features which distinguish it from other clauses. In order to have a complete chart comprising all possible case frames in Persian, a number of clauses was provided from my own intuition. Predicates as main elements carrying meaning within predications were shown to project features onto the nominals in their collocation. As noted in the discussions Of the Persian case frames (see Chapter 3), verbs are unique in the sense that each predicate may have a verb phrase as its surface representation, which may consist of one or two constituents. Verb forms may either be single--containing one verbal element only--or 'compound'-- a stable verbal element and a substitutable pre—verbal element. Using Longacre's analysis of case, I have shown that each predicate must be marked with features which specify the particular set of cases which occur in juxta— position to it. Furthermore, following Longacre, I classified parti- cipants at the predication level, adding two more roles to those which Longacre had previously devised for his analysis of English. The case labels had already been used by Pike and Pike (1977), and Sullivan (1980). Nevertheless, client (Sullivan's terminology), as used in this dissertation, was given a completely new definition to be compatible with the description of Persian case 281 frames. Moreover, examples were provided for action- process equative case frames, which Pike (1977) and Longacre (1976) had not discussed before. Based on the description provided for the Persian case frames, tagmemic formulas were arranged so that the deep-surface relationships of predications could better be demonstrated. An integrated tagmemic formula (see Chapter 3) was devised to make a conflation of those tag- memes whose occurrences are required in all types of clauses. To this end, we first noted that in the five sets of surface clauses analyzed, certain constituents were redundantly repeated. Constituents of this type were integrated into a general formula in order to simplify the formulas in such a way that a general picture of the deep-surface relationship Of the clauses in the language could be portrayed. This study primarily endeavored to answer what the deep cases and case frames in Persian are. In addition, it attempted to determine whether the selectional features attributed to the Persian verb cores lead to divergencies from English in surface morphological differences; and what points of conflicts exist between the two case systems of Persian and English. The analysis Of the Perisan case system (Chapter 3), and the comparison made between the case systems of English 282 and Persian (Chapter 5) showed that: 1. English and Persian show similarities in their underlying relationships. In other words, both systems are underlyingly characterized with similar features and are accompanied with similar arguments. However, what differentiates the two languages is the degree of surface interpolation. It was also revealed that the two languages differ in the sense that client and identifier as deep cases are required for the description of Persian but not for English. 2. Both English and Persian assign similar features to case frames in different predication groups. However, where one language represents predicates with different features in identical morphological forms, the other. superficially employs different forms. This is not always the case, since, in certain instances, both languages have different forms on the surface. For instance, English allows both experiencer and agent in collocation with the predicate smell. In Persian, SMELL surfaces as bu sanidan with the former, and buidan with the latter. 3. With the dative and accusative being marked in the surface structure of Persian, Persian provides greater flexibility in the ordering of clausal constituents than English. We showed (Chapter 5) that what English does with the word order, Persian does with case markers. In other words, a different arrangement of clausal elements in Persian does not distort meaning as much as it does in 283 English. Moreover, we showed that where English predications in groups G' and H' freely allow a collocation of path, source, and goal in the same simple predication, Persian structure tends to make a choice of coupling and embedding constructions. Certain rare concepts (see 5.0) were shown to be absent from English. 6.2. Limitations In the process of collecting the data, certain problems were encountered which may have affected the study and the analysis: 1. not many books were available, and the only novel which was available was the one used as a source of the data in this study. Other sources-- including plays, newspapers, and political articles-- were excluded since they could have affected the nature of the analysis otherwise. 2. the original intention was to share impressions concerning ambiguities and the areas of conflicts with other native speakers of Persian, except for two informants who regularly assisted in discussing the indeterminincies of sentence meanings, I was not able to enlist the aid of native speakers. 3. had a Persian novel and its translation into English by a professional translator been available, a 284 better analysis of clauses in the two languages involved may have been produced. 4. some of the articles on Persian verb systems referred to in Windfuhr (1979), and Moyne's dissertation on the Persian verb system as well were not available. It is possible that work similar to that presented in this dissertation has been conducted by Russian or German Iranist linguists. 6.3. The Contribution of This Study Aside from the two works carried out by Bashiri (see 4.4.3) and Palmer (see 4.4.1), the analysis of case frames in Persian as presented here appears to be the only avail- able detailed descriptive analysis Of predication in Persian. Palmer (1971) is a limited study of the syntax of the "ezafe" construction in Persian. In his analysis he attempts to account for the "ezafe" construction within the theoretical framework of Fillmore (1968). Bashiri's own theoretical framework is deveIOped in his work on case relationship in general, with Persian serving as a metalanguage to present examples which demonstrate these relationships (see Chapter 4, ft. 12). His case analysis is considerably different from the present one. In the present work, the major concern has been to focus on cases and case-predicate relationships in Persian. 285 Although English has been used to illustrate contrast, the aim has been to explicate the semantico-syntactic relation- ships in Persian clauses. This is the first attempt to incorporate features into the categorization of predica- tions in Persian.1 In the analysis of case frames presented here, certain predicates have been alluded to which not only determine the nature of roles attributed to the pre- dication participants but also the markers which are adjacent to the nominals on the surface. Although the conclusions in this respect are tentative because of the limitation of the data, we believe that they will be substantiated when the data is expanded and more examples are incorporated into the analysis. A classification Of predicates based on the features physical, experiential, directed, completable, and so on, though relatively old and well established for English may be a new development in Persian linguistics. Although there are almost certainly omissions, if the analysis presented here is deveIOped, taking into account other texts and materials, more regularities may be found and many questions answered which remained unresolved here. Finally, universal characteristics have been claimed to exist in any language (Falk 1978). In order to determine the principles shared by all or almost all languages, grammars of different languages should be first provided. 286 As Falk states, "...the search for universals of language and the writing of particular grammars are interrelated aspects of modern linguistics" (18). The present work may be a small step towards this goal. 6.4. Suggestions for Further Studies Suggestions for further studies take the forms of some primary but fundamental questions. To appreciate each one, the researcher needs to set up certain hypotheses and test them systematically, empirically, and accurately: 1. We noticed (see Table of Verb Entry Distribution p. 94 ) that the bulk of verb forms used in the construction of the indicative clauses in Daijan Napeleon were related to physical case frames. Is this true of all styles of writings? 2. Is the distribution of case frames at the para- graph and discourse level idiosyncratic? In other words, are different case frames selected willy-nilly and scattered idiosyncratically? Or are there generalizations that can be made? 3. From a sociolinguistic point of view, do all characters in the novel use all types of predica- tions and case frames equally? 4. What strategies does a good translator of a literary text use to superficially represent the divergencies in one language versus another? 287 5. Are the restrictions that the present work puts on notional categories based on predicate features verified if the data is expanded beyond the scOpe of the book used as a source of data? 6. Are there any differences in the selection of case frames between formal and casual writing? 6.5. Conclusion In this chapter, the major points discussed in the present work have been summarized. The contributions this work has presented were enumerated, and limitations which may have led to certain omissions were discussed. A set of questions relevant to the nature of case frame relationships was presented to suggest the possibility of extending the present or similar analyses to other works. Notes to Chapter Six 1As mentioned earlier, Russian and German scholars might have conducted similar studies. Case frames, as illustrated in this dissertation, may be significantly helpful in language teaching. Since differ- ent languages apply different strategies in representing deep-surface relationships, the teachers' comparisons of case systems in the languages involved and how surface structure and deep structure are related, can be effective in syllabus writing, material designing, and the methods of language teaching. 288 APPENDIX I SOME VERB ENTRIES IN PERSIAN In this section, we present a classification of some of the verbs which occurred in the data. The purpose is to show that a dictionary of a language needs to be more than a word list. A dictionary should represent the ways in which the vocabulary terms combine in the context. However, a complete list of verbs in Persian is beyond the scope of the present work; and the verbs presented here partially reflect the totality of simple and compound verbs in Persian. Since listing verbs alphabetically in Arabic and Roman scripts, and providing two sets of verb forms in two different a1- phabetical orders would be redundant, verbs will be ordered phonetically. Thus, verbs beginning with vowel sounds come first, and are followed by consonantal sounds. The order of verb entries will be as follows: i,e,a,§,o,u,p,b,t,d,k,g,q,m,n,l,r,f,v,s,z,§,z,x,h, 5.3” In the classification of verb entries, the following information is given for each verb: 1) the phonological form of the verb (Row 1). If the verb form appears in two different parameters, numbers will be assigned to the extended forms in order to differentiate them from the unmarked form. 289 290 2) the case frame of the verb, that is, the array of cases which obligatorily or optionally concur with the predicate as the core element of the case frame configuration will be presented (Row 1). Cases which are presented in parentheses are optional ones and not inherent to the verb valence. 3) the closest possible meaning in English (Row 2). 4) prepositions usually associated with each particular case (Row 2). The unmarked cases will not be repeated in this row. 5) typical example to illustrate the occurrences of cases in juxtaposition with the predicates (Row 3). Morpheme- to-morpheme translations as well as a close free translation will follow. Besides the conventional notations used in the abbreviation chart, Ex y j is used to demonstrate the part-whole relation- ships which act together in a certain way. VERB ENTRIES isténdan E A— A (L) to stand somi dar moqabel-e yek Edam-e mosamam istade-id you in front-E one man-E determined stand-you 'You are standing in front of a serious man' ist'adan2 = to stop nim sa'at ba'd doroske-e jelo-ye dar istad half hour later carriage-a front=E gate stopped 'Half an hour later, a carriage stOpped by the gate' 291 edame yaftan E {P-Phys‘} P to continue « estentaq edame yaft interrogation continuation found 'The interrogation continued' edrar kardan E A-Phys A (I) (inst) to urinate I: az Siyamak az tars edrfir kard Siyamak from fear urinated 'Siyamak wetted his pants in fear' eftetah Eodan E {P—Phys P to be inaugurated } ridiyo eftetah sode bud radio inauguration become was 'The radio station had started operating' ehsas kardan D'fAP-Experl complet A/E (G)R 1~directedj to feel G: nesbatbe Eqajan nesbatbe napeleon ehsEs—e ehteram ne-mi-kard Aqajan towards Napeleon feeling-E respect neg-did 'Aqajan had no feeling of respect towards Napeleon' ejri kardan E A- A R (S) complet to perform dastur-e u-ri ejri kard order-E him carried out 'He carried out his order' entezar kesidan E {A- A G to wait for someone man entezar-e to-ra mi-kesid-am I waiting-E you prog-pulled-I 'I was waiting for you' 292 enteqam gereftan C {AP-Exper} A E to take revenge E: az- tabi'at az daijan enteqam mi-gereft Nature from Daijan revenge IND-took 'Nature was taking its revenge on Daijan' etefaqjoftadan E {P-Phys P to happen complet } vaqe'e-ye ta'asofangizi etefaq oftad event-E regrettable happen fell 'A regrettable event occurred' andaxtan G'{A—Phys } A/S (path) G to throw Path: az- G: be- ketEb-ri az panjere be birun andaxt-am book-Ach from window to out threw-I 'I threw the book out of the window' anda'xtanl G {AP-Phys } A P L locative to put P: -ra ‘ L: ru- ma-ra andaxt ruye dug-a; we-Ach put on shoulder-his 'He put me on his shoulder' .. v ’ ' asabani sodan E P-Exper E to get angry ' } samsalimirza dist asabani mi-sod Samsalimirza prog angry prog-became 'Samsalimirza was getting angry' asar kardan C {AP-Exper}.A(RS)A E to influence/affect E: be- hayajan—e asdolamirza be hame asar kard excitement—E Asdolamirza to all affect did 'Asdolamirza's excitement influenced everyone' 293 avaz kardan E AP-Phys A P to change belaxare mozu'-r5 avaz kard-am finally matter-Ach changed-I 'Finally I changed the matter' az dast dadan H P-Phys P to lose possession man esq-am—ra az dast dad—am I love-my-Ach from hand gave-I 'I lost my love' 3b dadan E AP-Phys A P to water u daét gol-ha—ta ab mi-dad he prog flower-pl water prog-gave 'He was watering the flowers' adat dastan C S-Exper ‘ E R {complet ‘} to be accustomed to R: be- be nasayeh-e daijin adat dist—am to advise—E Daijan accustomed had-I 'I was accustomed to Daijan's advising' _ - , _ . —— . 2.52:: p s to come G: be/betarafe daijan az otaq birun Emad Daijan from room out came 'Daijan left the room' Eram kardan C{AP-Exper tR Sf E to cool off javab-e madar-am Bram-am kard answer-E mother-my calm—me did 'My mother's answer cooled me Off/calmed me' 294 aram sodan E P-Exper E to get calm qiyEfe-as Eram sode bud face-his calm become was 'He seemed to have been cooled off' aseqbudan D ’S-Exper E G to be in love with { daijan aseq-e nEpeleon bud Daijan lover-E Napeleon was 'Daijan loved Napeleon' aseq sodan D P-Exper E G to love/to fall in love with doxtar az tah-e qalb be aspirin aseq sode bud girl from bottom-E heart to Aspiran lover become was 'The girl had deeply fallen in love with Aspiran' avardan G' AP-Phys A P motion to bring v... V— .. masqasem cay avard Masqasem tea brought 'Masqasem brought (some) tea' avaz xandan D'rAP-Expef A/S E to sing 1 zarbgir avaz mi-xand drummer song prog-sang 'The drummer was singing' oftadan G' P-Phys ‘ P s G to fall [motion daijan az taxt be zamin Oftad Daijan from bed to ground fell 'Daijan fell off the bed' 295 oftadan1 to remember be yEd—e desm-hE-ye leyli oftad—am to memory-E eye-pl-E Leyli fell-I 'I remembered Leyli's eyes' pir kardan to age someone to marE pir kard-i you me Old did-you 'You aged me' peydi sodan to come to light asar-i az dozd peyda na-god C {P—Exper E {AP-Phys} E {P-Phys } sign-a from burglar appeared reg—became 'There was no sign of burglary' peyvastan to join masqasem be saf-e soja'an peyvast {A-Exper "Masqasem was honored to be among the heroes' panah bordan to shelter G' A- (Inst) v - - - az tars-e soma be qasab panah borde ast from fear-E you to butcher chelter taken is 'He has sheltered the butcher fearing you' paridan to jump vaqti mara did as ja parid when me saw from place jumped 'He jumped up when he saw me' G'{A-Phys } AP A/E G A G (I) A/P 296 paridan1 E P-Phys P(S)(I) to wake up (liti to jump) (Inst) az faryad-e u az xab parid-am from shouting-E him from sleep jumped-I 'His shouting woke me' partab kardan/part kardan G' AP-Phys A/S p to throw motion dustali giée-ye dava-rE partab kard Dustali bottle-E medicine threw 'Dustali threw the bottle of medicine' pasiman sodan C P-Exper E to feel regret dustalixan pasimin sod Dustalixan regretful became 'Dustalixan felt regret' paziroftan H {A-Phys }A/G P 5 to accept \ I hedye-ra az u paziroft-am gift-Ach from him accepted-I 'I accepted his/her gift' paziroftan1 C {A-Exper }A/E R to accept vaqiyat-e talXtra paziroft—am reality-E bitter accepted—I 'I accepted the sad reality' pare kardan E AP-Phys AP to tear kaqaz—ri pare kard paper-Ach torn did 'He tore the paper' 297 porsidan D' AP-Exper A R S Directed to ask Complet mi-xah-am Eiz-i az soma be-pors-am pres-want-I thing-a from you S-ask-I 'I want to ask you about something' bidar sodan C P-Exper E(S)(I) (Inst) to wake up S: az I: ba/az sobh-e zud az xab bidar god-am morning-E early from sleeping awake became-I 'I woke up early' birun kardan G' AP-Phys A PS(I) Motion to kick out (Inst) ’ - - . - - ...'- soma bayad mara az 1n xane ba ajan you must me from this house with police birun kon-id out do-you 'Only the police might force me from this house' band amadan E {P—Phys } P to be blocked sedE-yam band Emade bud voice-my blocked was 'I couldn't speak' bar dastan G' AP-Phys AP(S) Motion to take off S: az— asdoli eynak—e dudi-ye ura az gegmag Asdola glasses-E dark-E him from his eye bardast took 'Asdola took his dark glasses off' 298 baragofte budan C S—Exper 1E(I) to be restless/upset (Inst) J daijan saxt barasofte bud Daijan hard restless was 'Daijan was reStless/upset' bargastan G'{ - }AS Motion to return A: az daijan az mamuriyat-e sahrestEn-ha Daijan from mission-E city-pl bargaste bud returned was 'Daijan had returned from a mission to several other cities' barga's’tanl D A-Exper A/E G Directed to direct towards—- G: betarafe hame betarafe u barga§tand all towards he directed (their views) 'Everyone looked at him' bastan E AP-Phys ‘LAP to close dar-rE post-e sar-e va'ez bast-am door-Ach behind-E head-E preacher closed-I 'I closed the door when the preacher left' bil; bordan E A-Phys ‘}AP R(S) P to raise (e.g., to raise the voice) { J [ J kam kam sedE-yas-ra bala mi-bord little little voice-his- up prog-took 'He was gradually raising his voice' baz mandan E IS-Phys 'LP(I) (Inst) J to remain---(e.g., 'open') I: az- dahan-e hame az tahjob biz mind mouth-E all from wonder Open remained ' . . Everyon's mouth remained Open in wonder' 299 bordan G AP-Phys I A P G Motion I to take G: be mara bord t5 sangar-e xod-eman me took till trench-E self-our 'He took me to my own trench' bozorg kardan E AP-Phys AP to bring up doxtar bozorg kard-am mesle yek daughter big did-I like one daste-ye gol bunch-E flower 'lit. I brought up a daughter who is like a bouquet of flowers budan G S-Phys PL Locative to be L: tu/dar dustali dar Elam-e ro’b—o—vahsat-e xodag bud Dustali in world-E fear-and-fear-E himself was 'Dustali was in his own world of fear' budan1 1:{s-Phys } P Id. to be--- (e.g. something) masqasem model-e kuEek-e saxsiyat-e Masqasem model-E small-E character-E daijan bud Daijan was 'Masqasem was similar to Daijan in character' to punish Inst tanbih kardan C{AP-Exper} A E I madar-am mara tanib kard mother-my me punishment did 'My mother punished me' 300 tarsidan C {P-Exper ‘}E(I) (Inst) to fear I : az man az in damame mi-tars—am I from this demon pres-fear-I 'I am scared of this demon' tasmim gereftan C ’A-Exper A/E R to decide iComplet } pedar-at tasmim gerefte hame-ra davat kon-ad father-you decision taken all invitation does-he 'Your father has decided to invite everyone' v tasar zadan C IAP-Exper lA/S E(I) L(Inst) ‘1 to shout threats E: be daijan be u tasar mi-zad Daijan to him shouted 'Daijan used to shout threats at him' tagib kardan G' AP-Phys A P G to follow {Motion. } dosman-an-ra ta marz ta'qib kard—im enemy-pl- untill border followed-we 'We followed enemies to the border' tahif kardan D' AP-Exper A/S R E to tell (somethin) E:{ baraye :} baraye bade-hi qese-ha-ye ajibi tarif for child-pl story-pl-E strange mi-kard told 'He/she used to tell them strange stories' tirif kardanl D AP-Exper] A/G/E Directed to praise (oneself) (Reflexive) hamise az xodas tarif mi-kard always from himself praised 'He always praised himself' 301 didan D' P-Exper ER to see Complet man hargez yek ageq na-dide bud-am I never one lover neg—seen was-I 'I had never seen a person in love' dexalat kardan E ’A- A(G) to interfer { } faroxlaqa be hame-ye kEr-ha dexalat Farroxlaqa to all—E work-pl interference mi-kard prog-did 'Farroxlaqa interfered in all affairs' darax‘éidan E {A- A to shine xorsid mi-daraxsid sun IND-shine 'The sun was shining' davidan E {A- AG to run badge-hi betarafe dar-e bag david-and child-pl towards door-E garden ran—they 'The children ran towards the garden gate' davat kardan E ’A- AC aqajan sirali-ra davat kard { ‘} Aqajan Sirali invitation did 'Aqajan invited Sirali' dadan H AP-Phys ] A/S P G Possess Directedf to give G: be ' nim-i az bastani-ye xod-r5 be leyli did-am half-a from ice-cream self to Leyli gave-I 'I gave Leyli half of my ice-cream' 302 danestan C' S-Exper ER to know Complet man axlaq-e soma-r5 mi-dan-am I character-E you pres-know-I 'I know about your character' danestanl D {A-Exper} A/E RS to realize/recognize S: az ede’i in seda-ra az manga ensani some this sound from origin human danest-and knew-they 'Some believed that that sound came from a human source' dagtan H S-Phys GP to possess/to have possess dEijan haftir-i dist Daijan revolver-a had 'Daijan had a revolver' dur kardan G'IAP-Phys ‘ A P s to separate L j u-ra az zan-o-baEe-as dur kard-and him from wife-and-child-his far did-they 'He was forece to leave his family' to pull Motion kesidan G {AP-Phys A P(S)(G) leyli dast-e xod-r5 az labeléye angostan-e Leyli hand-E self from between fingers-E man birun kesid my out pulled 'Leyli took her hand from mine' \ kandan E AP-Phys A P S to pick { } gol-i az deraxt-ikand 0 be man did flower-a from tree-a picked and to me gave 'He picked a flower and gave it to me' 303 kastan c {AP-Phys A P L to plant pirsal dar mazra'e kalam kist-am last year in field cabbage planted-I 'I planted cabbage in the field last year' k5vidan E {'- AG to search (for something) . zan-as jib-hE-ye u-rE mi-kEv-ad wife-his pocked-pl-E his pres-search-she 'His wife looks into his pockets' kostan E AP-Phys A P I to kill man in hendi-ra be dast-e xodam mi-kog-am I this Indian-Ach to hand-E myself pres—kill—I 'I'll kill this Indian with my own hands' \ git kardan E P-Phys P(I) (Ins) to stick I: az sedi dar dahEn-e dustali az voice in mouth-E Dustali from fart-e xasm gir karde bud excessive anger stuck was 'Dustali was so angry his voice caught his throat' gerftan H' P-Phys ] G S P possess Motion j, to Obtain/to receive S: az sapur--- az dinesgih lisEns gerfte bud Sapur from university B.A. taken was 'Sapur had received his B.A. degree from the university' gereftan2 H A—Phys 1 A/G P S possessj to steal/to take S: az dustali esq-e mara az man gerefte bud Dustali love-E my from me taken was 'Dustali had taken/stolen my love from me' 304 gereftan3 E {AP-Phys L A P I Inst J to hold I: b5 gasEb-hi kard-r5 b5 dast-e rast butcher-pl knife with hand-E right mi-gir-and pres-hold—they 'Butchers hold the knife in their right hands' goftan D'{AP-Exper} A/S E R complet to say E: be Eqa be hasan gofte haq na-dar-ad Aga to Hassan said right neg-has-he pE-yag-ra az xane birun be-gozar-ad foot-his from house out S-put-be 'Aqa has prohibited Hassan from going out' gozastan G {AP-Phys:} A P L Locative to put L: ruye/tuye xod-am tuye qabr mi-gozEr-am-at self-my in grave pres-put-I-you 'I'll bury you' gozis’tanl E A- 7 AR to allow complet‘f u ne-mi-gozast kEr-i bo—kon-am he neg-pres-allow work-a S-do-I 'He didn't allow me to do anything' -V I gozastan2 E A— AR to leave complet u hie so'éli-ra bi-javab ne-mi-gozast he no question without-answer neg-pres-leave 'He left no question unanswered' 305 gazastan3 G P—Phy P Path Motion to pass by/to occur Path: az/dar ---hatt§ fekr-e yek zan-e digar ham even thought-E one woman-E other also dar maqz-am na-gozast in mind-my neg—passed 'I didn't even think of any other woman' qat' kardan E AP-Phys A P I to cut Inst u rige-ye deraxt-r5 qat' kard he root-E tree cut did 'He cut the root of the tree' qéd kardan E AP-Phys A P I to cut/to slice {Inst } hendavane-rE b5 anu qéd kard-am water melon with knife cut 'I sliced the watermelon with a knife' ngidan H SP-Phys lA/G P Possess to grab Motion J ketEb-ri az dast-e u qipid-am book-Ach from hand-E him grabbed-I 'I snatched the book from him' qol dadan D AP-Exper A E R to promise complet be u qu-e komak did-am to him promise-E help gave-I 'I promised to help him' qurt dadan E A- AR to swallow Eb-e gelu-rE qurt did water-E throat swallowed 'He swallowed his saliva' 306 I mahv sodan E P-Phys 1 PS ’-—J\—_ to disappear - — V V asar-e tars az surat-as mahv sod signe-E fear from face-his disappeared 'The signs of fear disappeared from his face' man' kardan E {A- A G R mara az xordan man' karde bud—and us from eating prohibited done were-they 'They had prohibited us from eating' maxfi kardan G {AP-Phys] A P L Locative? name—r5 dar guse—i maxfi kard-am ‘ letter-Ach in corner-a hidden did-I 'I hid the letter somewhere' mandan G A- L A to remain modati dar otEq-e entezar mind-am sometime in room—E waiting remained-I 'I waited in the waiting room for a while' 1 mandan G S-Phys PL to stay Locative atr-e leyli---be dast-ha-yam mimEnd perfume-E Leyli to hand-pl-my remained (lit. Leyli's perfume would remain on my hands) 2 leyli saket mind L j' Leyli silent remained ---m5ndan E [s-Phys ‘l P 'Leyli remained silent' momken budan Eh{S—Phys \k P to be possible hame 312 momken bud every thing possible was 'Everything was possible' 307 negah kardan D {h-Exper :}A/E G to look at-—- G: be faramarz be man negah mi-kard Faramarz to me looked 'Faramarz was looking at me' nesastan G{S-Phys }PL Locative to sit L: be araq be pigani-ye déijan nesaste bud sweat to forehead-E Daijan sat was 'Daijan's forehead was wet with sweat' nesindan G AP-Phys j A PL to sit (tr.) '{Locative’ bade-r5 ruye sandali nesind—am child-Ach on chair sat-I 'I put the child in his chair' nesan dadan D AP-Exper \M E G Instrumentf to show E: be b5 angost-e xod aks-e mara be u nesan mi-did with finger-E self picture me to him showed 'He was showing my picture to him' neveétan E A- } AR(G) to write 0: be/baraye/vaseye name-i be dust-am nevest-am letter-a to friend-my wrote-I 'I sent a letter to my friend' nafas kesidan E A- A to inhale nafas-i kesid-am breath-a pulled-I 'I inhaled' 308 flqfiv bastan E {P-Phys }P L to appear J labxand-i ruye lab-hE-yas naq; bast smile-a on lip—pl-his appeared 'A smile appeared on his lips' n'arihat budan C {S-Exper } E [:1] to be anxious baraye pedar-am narihat-am for father-my anxious-I am 'I am anxious for my father' lisidan E {AP-Phys } AP to lick boqub-e qazE-ye xod-r5 lisid plate-E food-E self licked 'He licked the plate of his food clean' \ lamidan G A- A :/ to lie down Locative diijin ruye taxt-as lamide bud Daijan on bed-his lying was 'Daijan was lying on his bed' laqzidan E P-Phys } P to slip pE-yas laqzid o oftid foot-his slipped and fell 'He slipped and fell down' larzidan B P-Amb 1 E(I) Exper to shiver \(Inst) J az sarma mi-larzid from cold prog-shivered 'He was shivering of cold' 309 larzidan1 to shake/to tremble tamEm-e badan-am larzid all-E body trembled 'My whole body started shaking' rixtan to pour golule az in taraf o in taraf mi—rixt E G{ P-Phys (Inst) P-Phys Motion bullet from this side and that side prog-poured 'Bullets were raining from everywhere' rixtan1 to enter/to rush into dozd-ha rixt-and tuye xeyme-ye m5 thief-pl poured-they in tent-E our 'The thieves attacked our tents' resEndan to get to name-ra be u resand-am letter-Ach to him got to 'I delivered the letter to him' resandan1 to reach (tr.) xod-r5 be panéhgih resEnd-am self to shelter reached (tr)-I 'I managed to reach to the shelter' rad kardan to reject pisnahEd-e u-ré rad kard-and suggestion-E him refused did—they 'They rejected his suggestion' I G{ ”I A... Motion AP-Phys Motion A-Phys Motion \_\,_z } P(I) A/P G 310 \ rad godan C {P-Exper L E to fail dar emtehEn rad god in exam failed 'He failed the exam' V rad sodan 1 G rA—Phys to pass by 1Motion nokar-e m3 zanbil be dast rad god servant-E our basket to hand passed became 'Our servant carrying the basket passed by' raftan E [A- to go L u dist mi-raft--- he prog prog-went 'He was going---' raftanl G A-Phys to flow Motion ab az xane-ye m5 be xane-ye water from house-E our to house-E EnhE mi-raft them prog-went 'The water flowed from our house to their house' rasidan G {p- to arrive S: az dustalixan az rah rasid Dustalixan from road arrived 'Dustalixan entered (the room)’ randan GJ’AP-Phys to send away L u-rE az xod-emin rind-im him from self-our sent away-we 'We ignored him/sent him away' 1 ‘ A/P Path h...\_..__! \ L A/P S G \ }, A P(S) 311 fekr kardan to think modati tulini fekr kard-am sometime long thought did-I 'I thought for a long time' fesordan = fegar didan to press dast-e leyli-r5 dar dast-am fesord-am hand-E Leyli in hand-my pressed-I 'I squeezed Leyli's hand in mine' fahmidan to understand m5 ne-mi-fahm-im we neg-pres-understand-we 'We/I don't understand' fahmidan1 to feel (lit. v - 0 man dard-e soma-ra m1-fahm-am to understand) I pain—E you pres-understand-I 'I feel/sense your pain' farapgereftan to cover 3b hame ji-ri fari gereft water every place covered 'Water covered everywhere' - V faramus kardan (sOmething) to forget man qam-e xod-r5 faramus kard-am I grief-E self forgot-I 'I forgot my own grief' c {A- A/E E {AP-Phys AP C S-Exper E(R) complet D' P-Exper ER complet E({A-Phys } AC G [P-Exper ‘ ER Lcomplet 312 farEr kardan E {Ar ) A(S) to run away 8: az gorbe-hi farér kard-and cat-p1 escape did-they 'The cats ran away' faryfid ke§idan E‘{Ar ‘} A/S to shout déijan faryEd kesid Daijan shout pulled 'Daijan shouted' Vahsat kardan C P-Exper E(I) (Inst) to be filled with terror I: az hame—ye hazerin vahsat karde bud-and all-E audience fear done were-they 'All the people present were terrified' vazidan E {P-Phys P bEd-e molEyemi mi-vazid wind-E gentle prog-blew 'A gentle breeze was blowing' sili xordan C P-Exper E to be slapped on the face sili-ye mohkami xord-am slap-E hard ate-I 'I was badly slapped on the face' sa'y kardan E A- AR complet sa'y kard-am fekr na—kon-am attempt did-I thinking neg-doeI 'I tried not to think' sEbet kardan E IA- 1 AR to prove Lcompletj man in-ri sEbet mi-kon-am I this prove pres—do-I 'I'll prove it' 313 saxtan E A- ‘ AR to build complet % pedar-as haft emirat sixte bud ’ father-his seven building built was 'His father had built seven buildings' sohbat kardan D'{AP-Exper }A/S E to talk to someone E: b5 modati b5 u sohbat kard-am sometime with him speech did-I 'I talked to him for a while' sorx sodan C P-Exper EI Inst to be ashamed I: az - v az xejalat sorx sod-am from shame red became-I 'I was ashamed of myself' so'al kardan D'IAP-Exper 1A E/S R to ask Lcomplet ‘I esm-e mar; az man so'il kard name-E my from me asked 'He asked me my name' zadan C A-Exper A/E I Intention to hit Inst masqasem be sar-e xod zad Masqasem to head-E self hit 'Masqasem hit himself on the head' zadan1 C A-Exper A E I '—‘—*-—— Intention to pat 'Inst dast-i be §§ne-ye man zad hand-a to shoulder-E my hit 'He patted me on the shoulder' 314 1 zadan2 G.{AP-Phys I A P G to put azgige-ye atr-e midar-am be surat—am zad-am from bottle-E perfume-E mother-my to face-my put-I 'I put some perfume from my mother's scent bottle on my face' zadan3 E AP-Phys 1 A P I EB‘EEEEt Inst J u yek golule zad tuye del-ag he one bullet hit in heart—his 'He shot him in the heart with a bullet' zadan4 E [AP-Phys I A P I Wt Ilnst I uri be qasd-e kogtan zad-am ' him to purpose-E kill beat-I 'I badly beat him up' zadan5 E AP-Phys ‘ A P I to bite Inst ZqE-ri mar zade Aqa snake bitten 'Aqa has been bitten by a snake' indan E A- AR togive birth to complet qamar do bade inde bud Ghamar two child borne was 'Ghamar had given birth to two children' zan bordan E{A- 1 A to marry to ne-mi-tavEn-i zan be-bar-i you neg-pres—can-you woman S-take—you 'you cannot marry' 315 ganidan E P-Exper \ E R(S) to hear {complet } sedE-ye masqasem-r5 sanid~dm voice-E Masqasem heard 'I heard Masqasem' ‘s’ath dadan D'{AP-Exper}A R E complet to explain E: bariye u avKeb-e vahgatnEk-e egq-ri bariye man he consequences-E fearful-E love for me sarh dad explanation gave 'He told me of the consequences of love' genaxtan D {S-Experq) EC to know (someone) DirectedI man in jens-e xabis-rE sal-hE-st mi-genas-am I this material-E dirty year-pl-is pres-know-I 'I have known this dirty rat for many years' Eekastan E AP-Phys 1 A P I to break (tr. and intr.) Inst ‘1 -— .- V sar-e u-ra ba sang sekast head-E his with stone broke 'He cracked his skull with a rock' gostan E AP-Phys A P I to wash Inst lebEs-ha-ra gost-am clothes-pl- washed-I 'I washed the clothes' xire godan D' A-Exper 1 A/E G to gaze {Directedf Eqajin be noqte-ye durdasti xire gode bud Aqajan to point-E far gaze became was 'Aqajna was gazing at a point a long distance away' 316 xarEb kardan E AP-Phys AP(I) (Inst) to demolish I: be xane—r5 xarab kard-am house ruined did-I 'I demolished the house' xaragidan E {P-Phys P to scratch (tr. and intr.) pE-yag be sang xarfigid foot-his to stone scratched 'He scratched his foot on stone' xabidan E P-Phys P (I) to sleep (Inst) man digab az negarini t5 sobh na-beid-am I last night from anxiety until morning neg-slept—I 'In my anxiety, I couldn't sleep last night' xabindan E AP-Phys] AP (I) to send to sleep {(Ins) belaxare baEe-ra xabind-am finally child slept (tr.)-I 'Finally, the child fell asleep' xabandanl E {AP-Phy; A P I to shoot to death [Inst } b; golule se t5 az dozd-hE-ri xgband-and with bullet three from thiefepl sleop (tr.)-3rd pl zamin ground 'He shot three thieves with one bullet' x'a'béndan2 E A- to turn down bayad sar—o-sedE-ye u-rE be-beZn—im must noise-E his S-turn down-we 'We must make sure he's quite' 317 xandan E A- AR name-ye leyli—r5 xand-am {:complet letter—E Leyli read-I 'I read Leyli's letter' xandan1 E {A~ }AR E3_§?Edy complet farémarz ketzb mi-xand Faramarz book prog-read 'Faramarz was studying' xgstan D {S-Exper 1’ EG doxtar-e gohar mi-xast J girl - def. husband IND-wanted 'The girl was looking for a husband' xordan E A- AR to eat sobhéne-ri xord-am breakfast ate-I 'I ate my breakfast' xoék kardan E AP-Phys ) A P(I) (Inst) I to dry (tr.) I: be lebis-hE-ri xogk kard clothes dry did 'He dried the clothes' xogk godan E {P-Phys ‘} P to freeze (lit. to die) ~ xande ruye lab-h§-§ xogk sod smile on lip-pl-her dry became 'The smile froze on her lips' hekiyat kardan D' AP-Exper A/S RE complet to tell E: baraye - - — V . - dastan-e xarab sodan—e otomObIl-ra story-E ruined become-E car 318 hekayat kard story did 'He told us the story of how the car had broken down' \ hes kardan D' P-Exper ER to feel complet hame-ye alE'em-ri hes mi-kard-am all-E signs feeling prog-did-I 'I could feel all the sings of...’ I hads zadan C {A-Exper } AR to guess hads mi-zan—am mariz b5§ad guess pres—hit-I sick be 'I presume he's sick' haram kardan C AP-Exper] A/E R complet to prohibit (Reflexive) xib-ra bar xod-emin harEm kard-im sleep on self-our unlawful did-we 'We didn't permit ourselves to sleep' harf zadan D' AP-Exper A/S ER complet to speak E: ba asdolEmirzE b3 madar-am harf mi-zad Asdolamirza with mother-my speech pres-hit 'Asdolamirza was talking to my mother' hol didan G AP-Phys AP(G) Directed to push G: be magqfisem marE hol d5d Masqasem me pushed 'Masqasem pushed me' 319 degm duxtan D'{A-Exper } A/E G to gaze G: be Magqisem degm be qeyEi-ye qubEni Masqasem eye to shears-E gardening duxte bud sewn was 'Masqasem was looking at the garden shears' dasbidan E {AP-Phys} A PC to stick G: be gu§—am-r§ be dar-e salon dasbind-am car-my to door-E hall stuck-I 'I put my ear to the door' 65p kardan/godan E{P-Phys 1 P to print/to be printed J ketEb—hE-ye aseqine---kam 55p éode bud book-pl-E amorous little print become was 'Very few books about love had been printed' \ iest zadan G A-Phys I A/P(G) to jump (towards) {Motion J nEyeb betarafe u jest-i zad Nayeb towards him jump-a hit 'Nayeb jumped towares/at him' iangidan E{ A- } AG to fight man b5 u jangid-am I with him fought-I 'I fought against them' javab dEdan D AP-Exper) A/S ER complet to answer Directed man be hame-ye soElEt-e u javab dad-am I to all questions-E his answer gave-I 'I answered all his questions' 320 to chew Inst flavidan E {AP-Phys 7 API tanEb-ri b5 dandEn-ag javid rope-Ach with tooth-his ch 'He chewed the rope with his teeth' joda' kardan c {AP-Phys} A P 3 Motion to separate S: az EnhE—ri az ham jodE kard-and them from each other separation did-they 'They were separated' Xi be I; kardan G{AP-Physl A P to move Motion I Eqijén va magqisem taxt-r3 j; be I; kard-and Aqajan and Masqasem bed-Ach place to place did-they 'Aqajan and Masqasem moved the bed' “Jusidan E{P-Phys } P to boil samavar mi-jugid Samovar prog-boiled 'The water in the Samovar was boiling' _APPENDIX II DATA BASE There are three types of examples used in the analysis of Persian case frames (Chapter Three): 1. examples extracted from Daijan ngeleon by Iraj Pezeshkzad, 2. examples extracted from Daijan Napeleon but modified for the purpose of simplicity, and 3. examples from my own intuition. The numbers used in this appendix refer to the examples used in the context. In the following list, 2 stands for Pezeshkzad's Daijan Napeleon; M stands for 'modified examples'; and A stands for examples from my own intuition. D(A) means that the original sentence is from D but the subsentences are mine. SENTENCE SOURCE SENTENCE SOURCE l A 14 A 2 A 15 A 3 A 16 D 4 A 17 D 5 D 18 M 6 D 19 D 7 D 20 M 8 D 21 A 9 M 22 A 10 A 23 A 11 A 24 D 12 D 25 A 13 D 26 A 321 322 SENTENCE SOURCE SENTENCE SOURCE 27 D 72 M 28 A 73 M 29 D 74 D 30 A 75 D 31 A 76 D 32 A 77 D 33 A 78 M 34 A 79 A 35 A 80 D 36 A 81 D 37 D 82 D 38 D 83 D 39 M 84 D 40 A 85 D 41 A 86 D 42 A 87 A 43 D 88 D 44 D 89 D 45 A 90 D 46 A 91 D 47 M 92 M 48 A 93 M 49 A 94 D 50 A 95 D 51 A 96 D 52 M 97 M 53 M 98 A 54 D 99 A 55 A 100 A 56 D 101 D 57 D 102 A 58 D 103 A 59 D 104 A 60 D 105 D 61 A 106 A 62 A 107 A 63 D 108 A 64 D 109 D 65 D 110 D 66 D 111 D 67 D 112 D 68 D 113 D 69 D 114 D 70 D 115 M 71 D 117 A 323 SENTENCE SOURCE SENTENCE SOURCE 118 M 164 D 119 D 165 D 120 D 166 D 121 A 167 M 122 D 168 A 123 M 169 D 124 D 170 A 125 M 171 A 126 M 172 A 127 M 173 A 128 M 174 A 129 M 175 D 130 A 176 M 131 A 177 D 132 A 178 M 133 M 179 A 134 A 180 D 135 D 181 D 136 A 182 D 137 A 183 D 138 A 184 A 139 D 185 D 140 A 186 A 141 A 187 D 142 M 188 M 143 D 189 D 144 D 190 D(M) 145 D 191 D 146 D 192 M 147 D 193 D 148 A 194 D 149 A 195 D 150 D 196 D 151 D 197 D 152 A 198 M 153 A 199 A 154 M 200 D 155 D 201 M 156 D 202 A 157 D 203 A 158 D 204 D 159 M 205 A 160 D 206 A 161 D 207 A 162 D 208 A 163 D 209 A 324 SENTENCE SOURCE SENTENCE SOURCE 210 A 257 M 211 A 258 A 212 D 259 A 213 A 260 A 214 A 261 A 215 A 262 M 216 A 263 M 217 A 264 D 218 A 265 D 219 A 266 D 220 A 267 A 221 A 268 A 222 A 269 D 223 A 270 M 224 M 271 A 225 A 272 A 226 A 273 M 227 A 274 A 228 A 275 M 229 A 276 D 230 A 277 D 231 A 278 D 232 A 279 D 233 A 280 D 234 A 281 A 235 M 282 A 236 A 283 M 237 D 284 M 238 D 285 D 239 M 286 M 240 D 287 D 241 M 288 D 242 D 289 D 243 A 290 D(A) 244 M 291 A 245 A 292 D 246 A 293 D 247 M 294 D 248 A 295 A 249 D 296 D(A) 250 D 297 D 251 D 298 D 252 A 299 D 253 M 300 D 254 D 301 D 255 A 302 D 256 D 303 D 325 SENTENCE SOURCE SENTENCE SOURCE 304 D 348 A 305 A 349 A 306 M 350 D 307 D 351 A 308 M 352 A 309 D 353 M 310 D 354 D 311 D 355 A 312 D 356 A 313 D 357 M 314 M 358 A 315 M 359 M 316 M 360 A 317 A 361 A 318 A 362 PIKE 319 A 363 PIKE 320 A 364 PIKE 321 A 365 D 322 A 366 A 323 A 367 A 324 A 368 A 325 D 369 A 326 D 370 A 327 D 371 D 328 D 372 M 329 M 373 A 330 D 374 A 331 A 375 D(A) 332 D 376 M 333 D 377 A 334 D 378 M(A) 335 D 379 M(A) 336 M 380 A 337 D 381 A 338 D 382 M 339 D 383 A 340 M 384 A 341 M 385 M 342 D 386 A 343 M 387 M 344 A 388 A 345 M 389 A 346 D 390 A 326 SENTENCE SOURCE 391 392 393 394 395 33>33>> BIBLIOGRAPHY Ahsan, Abd-o-shakur, 1963. "Modern Persian Grammar." Pakistani Linguistics, Lahore, pp. 99-129. Amid, Hassan, 1971. Farhange Amid (Persian Dictionary). Tehran: Entesharate Javidan. Anderson, John M., 1971. The Grammar of Case: Towards a Localistic Theory. Cambridge: The University Press. . 1977. On Case Grammar. London: Groom Helen. Arberry, Arthur J., 1942. British Contributions to Persian Studies. London: Longmans Green. Baker, H.G., 1931. "Case in Some Earlier and Later English Grammars." Papers of the Michigan Academy of Science, Arts and Letters, 14, 525-35. Barroomand, A.A.K., 1963. The Persian-English Dictionary. Tehran: Pirooz Printing and Publications Institute. Bashiri, Iraj, 1972. To Be as the Origin Of Syntax: A Persian Framework. Ph.D. Dissertation. University of Michigan. BEteni, Mohammad Reza, 1967. Towsif-e ngtemSn-e Dasturi-ye ZabEn-e Farsi. (Description of the Structure of Persian). Tehran: Amir kabir. * Beekman, John, 1968. "Elliciting Vocabulary, Meaning, and Collocation." Notes on Translation, 29, 1-11. Blansitt, Edward, 1969. ”Sentence and Clause in Universal Grammar." Anthropological Linguistics, 12, 112-21. Bolinger, Dwight, 1975. Aspects of Language. New York: Harcourt Brace Jovanovich. Brend, Ruth M., 1972. "Tagmemic Theory: An Annotated Bib— liography." Journal of English Linguistics, 6, 1—16. 327 328 _____m_", 1974. Advances in Tagmemics. Amsterdam: North- Holland Publishing Company. Burling, Robins, 1970. Man's Many Voices: Language in its Cultural Setting. New York: Holt. Chafe, Wallace, 1970. Meaning, and the Structure of Language. Chicago: University of Chicago Press. . 1971. "Directionelity and Paraphrase." Language, 47, 1-260 Chomsky, Noam, 1975. Syntactic Structures. The Hague: Mouton. . 1965. Aspects of the Theory of Syntax. Cambridge, Mass.: MIT Press. Chomsky, Noam, 1976. Reflections on Language. New York: Pantheon Books. . 1979. Language and Responsibility. New York: Pantheon. . 1980. "On Binding." Linguistic Inquiry, 11, 1-46. Cook, Walter A., 1970. "Case Grammar: From Rules to Roles." Languages and Linguistic Working Papers, 1. Washington, D.C.: Georgetown Univ. Press, pp. 14-29. . 1971a. "Case Grammar as Deep Structure in Tagmemic Analysis." Languages and Linguistic WorkingyPapers, 2. Washington, D.C,: Georgetown Univ. Press, pp. 1-9. . 1971b. "Improvements in Case Grammar, 1970." Languages and Lingustic Working Papers, 2. Washington, D.C.: Georgetown Univ. Press, pp. 10-22. . l972a. "A Set of Postulates for Case Grammar Analysis." Languages and Linguistic Working Papers, 4, Washington, D.C.: Georgetown Univ. Press, pp. 35-49. . 1972b. "A Case Grammar Matrix." Languages and Linguistic Working Papers, 6. Washington, D.C.: Georgetown Univ. Press, pp. 15-47. . 1973. "Covert Case Roles." Langpages and Linguistic Working Papers, 7. Washington, D.C.: Georgetown Univ. Press, pp. 52-81. . 1979. Case Grammar: Development of the Matrix Model (1970-1978). Washington, D.C.: Georgetown Univ. Press. 329 Corder, S.P., 1968. "Double-Object Verbs in English." Studia Anglica Posnaniensia (SAP), 1, 15-18. Deibler, Ellis W., 1968. "Translating From Basic Structure." The Bible Translator, 19, 14—16. Dik, Simon C., 1980a. Studies in Functional Grammar. London: Academic Press. . 1980b. "Basic Principles and Application of Functional Grammar." Syntax and Semantics: Current Approaches to Syntax, 13, Ed. by Edith A. Moravcsik, and Jessica R. Wirth. New York: Academic Press, pp. 45-75. Dressler, Wolfgang U., 1970. "Towards a Semantic Deep Structure of Discourse Grammar." Papers from the Sixth Regional Meeting of the Chicago Linguistic Society, Chicago: University of Chicago, Dept. of Linguistics. pp. 202-209. Extiar, Mansur, 1967. "A Semantic Problem in Translation." Journal of the Regional Cultural Institute, 1, 61-67. Falk, Julia 3., 1978. Linguistics and LanguAge: A Survey of Basic Concepts and Implications (2nd edition). New York: John Wiley & Sons. Fillmore, C.J., 1966. "Deictic Categories and the Semantics of 'Come'." Foundations of Language, 2, 219-227. . 1967. "The Syntax of English Preverbs." Glossa, 1, 91-1250 . 1968a. "The Case for Case." Universals in Lin- guistic Theory, Ed. by Emmon Bach, and Robert T. Harms. New York: Holt, Rinehart and Winston, Inc., pp. 1-90. . 1968b. "Lexical Entries for Verbs." Foundation of Langgage, 4, 373-393. . 1969. "Types of Lexical Information." The Ohio State University Working Papers in Linguistics, 2, . 1971. "Some Problems for Case Grammar." Monograph Series on Languages and Linguistics, 24, pp. 35-55. . 1977. "The Case for Case Reopened." Syntax and Semantics, 8, Ed. by Peter Cole, and Jerrold Sadock (1977). New York: Academic Press, 59-82. 330 Firbas, J., 1966. "Non-thematic Subjects in Contemporary English." Travaux Linguistiques de Prague, 2, 239-59. Forbes, Duncan, 1861. A Grammar of the Persian Language. London: W.H. Allen and CO. Fries, Charles C., 1952. The Structure of English: An Introduction to the Construction of English Sentences. New York: Harcourt, Brace and World. Greenberg, Joseph H., 1972. "Some Universals of Grammar with particular Reference to the Order of Meaningful Elements." A Reader in Historical and Comparative Linguistics (1972), Ed. by Allan R. Keiler. New York: Holt, Rinehart and Winston, Inc., pp. 306-337. Groot, A.W.de., 1957. "Subject-Predicate Analysis." Lingua, 6, 301-318. . 1966. "Classification of Cases and Uses of Cases." For Roman Jakobson, The Hague: Mouton, pp. 189-194. Halliday, M.A.K, 1967. "Notes on Transitivity and Theme in English." Journal of Linguistics, 3, 37-81; pp. 179-215. . 1968. "Notes on Transitivity and Theme in English." Journal of Linguistics, 4, 179-215. . 1975. "The Teacher Taught the Student English: An Essay in Applied Linguistics.: LACUS FORUM, 2, 344—49. Henderson, Michael M.T., 1978. "Modern Persian Verb Morphology." Journal of the American Oriental Sociegy, 98, 375-88. Hjemslev, L., 1953. La categorie de cas. Copenhagen: Aarhus Univ. Press. Hockett, Charles F., 1961. "Linguistic Elements and their Relations." Language, 37, 29-53. Homayun-Farrokh, A. 1959. Dastur-e J5me'-e ZabEn—e Firsi (A Comprehensive Grammar of Persian). Tehran: Elmi Publications. Hopper, P.J. and S.A. Thompson, 1980. "Transitivity in Grammar and Discourse." Language, 56, 251-299. Hudson, Williams T. 1936. A Short Grammar of Old Persian. Cardiff: The University of Wales Press. Jackendoff, Ray, 1972. Semantic Interpretation in Generative Grammar. Cambridge, Mass.: MIT Press. 331 Jazayery, Mohammad-Ali, 1966. "Western Influence in Con- temporary Persian: A general Review." BSOAS, 29, 79-96. . 1970. "Modern Persian Prose Literature." Journal of the American Oriental Society, 90, 257-65. Jones, Linda K., 1980. "A Synopsis of Tagmemics." Syntax and Semantics: Current Approaches to Syntax, 13, Ed. by Edith A. Wirth. New York: Academic Press. 1980, pp. 77-97. Jones, W., 1717. A Grammar of the Persian Language. England: The Scholar Press. Khanlari, Parviz Natel, 1973. TErix-e ZabEn-e Firsi (The History of Persian). I and II, Tehran: Iranian Cultural Foundation. Kuryowicz, Jerzy, 1950. "Aspect et temps dans l'historie du persan." Rocznik Orientalistyczny, 19, 531-42. Langacker, Ronald, 1972. "A Review of Wallace Chafe, Meaning. and the Structure Of Langauge." Language, 48, 134-161. Lazard, Gilbert, 1957. Grammaire du persan contemporarian. Paris: Librarie C. Klincksieck. Levy, Reuben, 1951. The Persian Language. London: Hutchinson. Lockwood, David C., 1972. Introduction to Stratificational Linguistics. New York: Harcourt, Brace, and Jovanovich. Longacre, Robert E., 1960. "String Constituent Analysis." Language, 36, 63-88. . 1970. "Hierarchy in Language." Method and Theoyy in Linguistics. Ed. by Paul Garvin. The Hague: Mouton. pp. 173-95. . 1973. "Some Implications of Deep and Surface Structure Analysis for Translation." Notes on Trans- lation, 45, 2-10. . 1976. An Anatomy of Speech Notions. Lisse, the Netherlands: The Peter De Ridder Press. Lyons, John, 1963. Structural Semantics. Oxford: Blackwell. . 1966. "Towards a 'Notional' Theory of the 'Parts of Speech'." Journal of Linguistics, 2, 209-36. 332 MacKinnon, Colin, 1977. "The New Persian Preverb BI-." Journal of the American Oriental Society,97.l, 8—26. Maclay, Howard, 1970. "Overview." Semantics: An Inter- discriplinary Reader in Philosophy, Linguistics and Psychology. Ed. by Jakobovits and Steinberg, Cambridge University Press, pp. 157-181. Magnusson, R., 1954. Studies in the Theory of the Parts of Speech. Lund: Gleerup. Makkai, Adam, 1972. Idiom Structure in English. The Hague: Mouton. Manuchehri, Parivash, 1974. "Explaining Problems of Iranian Students by Comparing English and Farsi Verb Forms." TESOL Quarterly, 8, 171-76. Marashi, Mehdi, 1970. The Persian Verb: A Partial Description for Pedagogical Purposes. Ph.D. Dissertation. University of Texas. . 1972. "Modal and Auxiliaries in Persian." Orbis, 21, 417-28. Matson, D.M., 1959. "Tagmemic Description of Agreement." Monograph Series on Languages and Linguistics, 20, pp. 168-76. Milanian, Hormoz, 1969. "Les monemes et les procedes de passage en persan: de la categorie verbale a la categorie nominale et vice versa." Word, 25, 214-227. Moyne, John Abel, 1971. "Reflexive and Emphatic." Language, 47, 141-63. . 1974. "The So-called Passive in Persian." Linguistic Inquiry, 5, 205-49. Newmeyer, Frederick J., 1980. Linguistic Theory in America: the First anrter-century of Transformational Generative Grammar. New York: Academic Press. Nilsen, Don L.F., 1971. "The Use of Case Grammar in Teaching English as a Foreign Language." TESOL Quarterly, 5, 293-99. Palmer, Adrain S., 1971. The 'Ezafe' Construction in Modern Persian. Ph.D. Dissertation. University of Michigan. 333 Palmer, F.R., 1962. "Grammatical Categories and their Exponents." Proceedings of the Ninth International Congress of Linguistics, pp. 338-45. . 1965. A Linggistic Study of the English Verb. London: Longmans. Paper, Herbert H., 1959. "Review of Gilbert Lazard, Grammaire dn_persan contemporarian (1957)." Journal of American Oriental Society, 79, 31-39. Peterson, David, 1974. Noun Phrase Specificity. Ph.D. Dissertation. Universtiy of Michigan. Pezeshkzad, Iraj, 1977. Daijan Napeleon. Tehran: Entesharate Safialishah. Phillott, D.C., 1919. Higher Persian Grammar. India: British Mission Press. Pike, Kenneth L., 1944. "Analysis of a Mixteco Text." International Journal of American Linguistics, 10, 113-138. . 1962. "Dimensions of Grammatical Constructions." Language, 38, 221-44. . 1967. Language in Relation to a Unified Theory of the Structure of Human Behavior (2nd Revised Edition). The Hague: Mouton. Pike, Kenneth L. and Evelyn G. Pike, 1977. Grammatical Analysis. Arlington, Texas: Summer Institute of Linguistics and the University of Texas at Arlington. Platt, John T., 1971. Grammatical Form and Grammatical Meaning: A Tagmemic View of Fillmore's Deep Structure Case Concepts. Amsterdam: North-Holland. Robins, R.H., 1952. "Noun and Verb in Universal Grammar." Language, 28, 289-98. . 1970. "The Development of the Word Class System of the European Grammatical Tradition." Foundations of Language, 2, 3-19. 1974. A Short History of Linguistics. Bloomington: Indiana University Press. Rosen, Friedrich, 1898. Modern Persian Colloquial Grammar. London: Luzac. 334 Saussure, F. De., 1916. Course de linguistique generale. English Translation by Wade Baskin, Course in General Linguistics. New York: Philosophical Library, 1959. Sheintuch, Gloria, 1976. "Periphrastic Verb Formation in Persian." Studies in the Linguistic Sciences, 5, 139-156. Starosta, Stanely, 1971. "Lexical Derivation in Case Grammar.‘ University of Hawaii Workinngapers in Linguistics, 3, 83-101. . 1976. "A Place for Case." Langggge Learning, 26, 1-36. Sullivan, William J., 1980. "Syntax and Semantics in Strat- ificational Theory." Syntax and Semantics: Current Approaches to Syntax, 13, Ed. by Edith A. Moravcsik, and Jessica R. Wirth, 1980. New York: Academic Press pp. 301-329. Tabaian, Hessam, 1979. "Persian Compound Verbs." Lingua, 47, 189-208. Windfuhr, Gernot L., 1979. Persian Grammar: History and State Of its Study. The Hague: Mouton. Yarmohammadi, Lotf-ol-lah, 1965. A Contrastive Study of Modern Persian. Ph.D. Dissertation. Indiana University. . 1970. "A Note on Contrastive Analysis." English Language Teaching, 25, 76-79.