flaky;

.n. fine
a

«a.

1. It;
.3 1.."

a...

k 12.;

3: X?»
F..
r

 

n«- p
,2;;

. at
“JP...
L

 

 

 

I K A P . . "-, nvl.M¢ I“.
... is u: u .. . . ; . 13.3.“:

This is to certify that the
thesis entitled

THE EFFECTS OF THE EXPLICIT TEACHING METHOD ON
LIAISON ACQUISITION BY L2 LEARNERS OF FRENCH

presented by

MICHEALE ELIZABETH MICOL

has been accepted towards fulfillment
of the requirements for the

MA. degree in Adult Language Learning

 

 

 

 

 

/'\t©m\"

 

mgr/DEL (121001

Date

MSU is an Afllnnafive Action/Equal Opportunity Institution

 

 

v. ON '-—-- '

1—- -_-'. r-

 

 

LIBRARY
Michigan State
University

 

 

PLACE IN RETURN BOX to remove this checkout from your record.
To AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE

DATE DUE

DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6/01 cJCIRC/DateDuepss-p. 15

THE EFFECTS OF THE EXPLICIT TEACHING METHOD ON LIAISON
ACQUISITION BY L2 LEARNERS OF FRENCH

By

Micheale Elizabeth Mice]

A THESIS

Submitted to
Michigan State University
In partial fulfillment of the requirements
For the degree of

MASTER OF ARTS

Department of Linguistics and Germanic,
Slavic, Asian, & African Languages

2004

ABSTRACT

THE EFFECTS OF THE EXPLICIT TEACHING METHOD ON LIAISON
ACQUISITION L2 LEARNERS OF FRENCH

By

Micheale Elizabeth Micol

The current study examines the effects of teaching method on liaison production by
second language learners of French. Three test groups were used to examine the effects
of explicit instruction, implicit input, and explicit instruction in addition to implicit input
on liaison production. The explicit instruction group consisted of eight students taking
French 330 Phonetics. The implicit group consisted of five students taking 300-level
French courses that did not include French 330 Phonetics. The implicit plus instruction
group consisted of students taking French 330 Phonetics concurrently with another 300-
level French course. The groups were tested in a pretest/posttest format and two raters
were used to determine the presence or non-presence of liaison consonants. The
independent variables used were instructional treatment (implicit, instruction, implicit +
instruction), reading style (paragraph, phrase list), and time (pretest, posttest). The
dependent variable is correct liaison production. Results suggest that the explicit
teaching method had no significant effect on correct liaison production by these learners.
In addition to exploring the explicit teaching method, consideration is given to the

characterization of liaison usage in these learners’ speech.

ACKNOWLEDGEMENTS

First and foremost I would like to thank Professor Anne Violin-Wigent for
chairing my committee and also for sharing her vast knowledge of the French Language.
Thank you very much for supporting and encouraging me throughout my graduate career.
Additionally, I would like to thank Professors Debra Hardison, Grover Hudson and
Cristina Schmitt for taking time to be on my committee and helping me progress as a
researcher. Many thanks to Professors Anna Norris, Severine Collignon and Ehsan
Ahmed for allowing me to use their classes as my volunteer base and also to Professor
Dennis Gilliland for his assistance with the statistics portion of this study.

I would like to acknowledge Jessica Stakoe, Jennifer Miller, and Kristin
Whitwam for their help in taping, as well as Isabelle Choquet and Gerard Beck for their
help in rating the data. I would like to show appreciation to all of the study participants
for taking time out of their busy schedules to help with this study. Finally, I would like to
say thanks to my loving husband, who sat up with me many nights, supporting and

encouraging me as I went through the process of researching and writing this thesis.

iii

TABLE OF CONTENTS

LIST OF TABLES .................................................................................. vi
CHAPTER ONE
INTRODUCTION .................................................................................... 1
CHAPTER TWO
BACKGROUND INFORMATION ................................................................. 3
2.1 Liaison ..................................................................................... 3
2.1.1 Phonology of Liaison ....................................................... 4
2.1.2 Phonetics and Orthography ................................................ 5
2.1.3 Liaison Consonants and Grammatical Categories ....................... 7
2.1.4 Prosody ...................................................................... 11
2.1.5 Syntax of Liaison ............................................................ 13
2.1.6 Contexts of Liaison .......................................................... 16
2.2 Teaching Liaison ........ ' ............................................................... 26
2.3 Explicit and Implicit Instruction ..................................................... 29
2.4 Conclusion ............................................................................. 30
CHAPTER THREE
EXPERIMENTAL DESIGN ...................................................................... 33
3.1 Research Questions and Hypotheses ................................................ 33
3.2 Method ................................................................................... 35
3.2.1 Experiment Variables ....................................................... 35
3.2.2 Participants .................................................................... 35
3.2.2.1 Implicit Group (IMP) .......................................... 36
3.2.2.2 Instructional Group (INS) ..................................... 36
3.2.2.3 Implicit + Instructional Group (IMPINS) ................... 37
3.2.3 Raters ......................................................................... 37
3.2.4 Test Medium ................................................................. 38
3.2.5 Questionnaire ................................................................. 41
3.2.6 Procedure ..................................................................... 41
CHAPTER FOUR
RESULTS AND DISCUSSION .................................................................. 44
4.1 Hypothesis One: Teaching Methods ................................................ 44
4.2 Hypothesis Two: Speech Style ...................................................... 45
4.3 Hypothesis Three: Obligatory versus Impossible Liaisons ...................... 46
4.4 Hypothesis Four: Sequences with Deterrniners .................................. 48
4.5 Hypothesis Five: Sequences with Verbs and Pronouns ......................... 48
4.6 Hypothesis Six: Set Phrases ......................................................... 49
4.7 Hypothesis Seven: Liaison Consonants ........................................... 50
4.8 Further Discussion .................................................................... 52

iv

CHAPTER FIVE

CONCLUSION ...................................................................................... 56
APPENDIX A

PARTICIPANT INFORMATION ............................................................... 59
APPENDIX B

EXPLICIT INPUT FROM TEXTBOOK ........................................................ 60
APPENDIX C

TEST FOR PARTICIPANTS ..................................................................... 69
APPENDIX D

RATER INSTRUCTIONS ........................................................................ 71
APPENDIX E

ANONYMOUS QUESTIONNAIRE ............................................................ 74
REFERENCES ................................................................................. . ..... 'T7
ENDNOTES ......................................................................................... 79

LIST OF TABLES

TABLE 2.1
ORTHOGRAPHIC/PHONETIC LIAISON CONSONANTS ................................ 7
TABLE 2.2

GRAMMATICAL CONTEXTS UTILIZING [z] ............................................... 3
TABLE23

GRAMMATICAL CONTEXTS UTILIZING [t] ................................................ 9
TABLE 2.4

GRAMMATICAL CONTEXTS UTILIZING [n] ............................................... 10
TABLE 2.5

RARE LIAISON CONSONANTS ................................................................ 11
TABLE 2.6

OBLIGATORY LIAISON LICENSING ........................................................ 20
TABLE 2.7

OPTIONAL LIAISON LICENSING ............................................................ 22
TABLE 2.8

IMPOSSIBLE LIAISON LICENSING ......................................................... 25
TABLE 2.9

OBLIGATORY AND IMPOSSIBLE LIAISON CONTEXTS .............................. 26
TABLEZJO

RESEARCHER AGREEMENT .................................................................. 31
TABLE3J

OBLIGATORY AND IMPOSSIBLE LIAISON CONTEXTS .............................. 39
TABLE 3.2

OBLIGATORY CONTEXT DATA SET ....................................................... 40
TABLE33

IMPOSSIBLE CONTEXT DATA SET ......................................................... 41
TABLE4J

CORRECT LIAISON PRODUCTION VERSUS TIME ..................................... 44

vi

TABLE 4.2

CORRECT LIAISON PRODUCTION ACCORDING TO SPEECH STYLE ............. 45
TABLE 4.3

OBLIGATORY AND IMPOSSIBLE PRODUCTION ........................................ 47
TABLE 4.4

DETERIVIINER—NOUN VERSUS DETERMINER-ADJECTIVE .......................... 48
TABLE 4.5

PERSONAL PRONOUN-VERB VERSUS VERB-PRONOUN SEQUENCES .......... 49
TABLE 4.6

OBLIGATORY VERSUS IMPOSSIBLE SET PHRASES ................................... 49
TABLE 4.7

LIAISON PRODUCTION ACCORDING TO LIAISON CONSONANT ................. 51
TABLE 4.8

RECOGNITION OF LIAISON ................................................................... 52
TABLE 4.9

PRODUCTION OF LIAISON .................................................................... 53
TABLE 4.10

SURVEY QUESTIONS 3, 8 ...................................................................... 54
TABLE 4.11

SURVEY QUESTIONS 4, 9 ...................................................................... 55
TABLE A. 1

PARTICIPANT INFORMATION ............................................................... 59
TABLE B. 1

OBLIGATORY LIAISON ........................................................................ 62
TABLE B.2

H-ASPIRE ........................................................................................... 63
TABLE B.3

IMPOSSIBLE LIAISON .......................................................................... 64
TABLE B.4

OPTIONAL LIAISON ............................................................................. 65

vii

TABLE 35
NUMERALS IN LIAISON ........................................................................ 66

viii

CHAPTER ONE
INTRODUCTION
Liaison is a feature of the French language that, succinctly put, causes a
characteristically latent word-final consonant to be realized phonetically in certain

linguistic environments. For example, one would pronounce ‘my cat’, man chat, as

[mile] but ‘my friend’ man ami as [mfmami]. This phenomenon can be loosely
compared to a/an alternation in English (T ranel 1987: 168). ‘A car’ is pronounced [okar]
but ‘an apple’ is pronounced [mp1].

Liaison usage is common in French speech and proper employment of it can
indicate a mastery of the language system (Battye et al. 2000:112). Thus, liaison usage is
important to second language (L2) learners who wish to be proficient in speaking French.
Additionally, liaison usage can also be a gauge for speech style since it tends to be
employed more frequently in formal situations. Since liaison is an integral part of the
French language, attention should be given to it in second language teaching and
learning. The present study investigates the effects of a specific teaching method on
liaison production by L2 learners, as well as general characteristics of liaison usage by
these same learners. In this study, two research questions will be explored.

1. How does an explicit teaching method influence liaison production in second
language speech?

An explicit teaching method is one in which the teacher presents the lesson to be learned
by explaining the phenomenon and the rules that govern it, usually in lecture format.
Learners are not necessarily required to use inductive reasoning and draw their own

. conclusions, but rather are supplied with the information they need to know at the onset

of the lesson. Exploring this question will provide evidence as to whether this method is
effective for teaching liaison.

1. How do (L2) learners of French employ liaison in speech?
Analyzing L2 speech data for liaison usage will provide information as to how learners
employ liaison in speech. This will in turn identify areas in which learners may need
clarification or practice in liaison.

Finding answers to the aforementioned questions will benefit both second
language teaching and second language acquisition research. By examining learners’
interlanguage speech, acquisition of liaison can be examined. Additionally, this
information will allow instructors to examine the effectiveness of an explicit teaching
method on liaison production and to identify areas in which L2 learners may be in need
of clarification and/or practice of concepts. Through the examination of the explicit
teaching method, instructors will be provided with useful information as to the
effectiveness Of such a method. By examining the interlanguage Speech of L2 Ieamers,
the acquisition of liaison can be studied and areas where learners may need help can be

identified.

CHAPTER TWO
BACKGROUND INFORMATION
2.1 Liaison
As discussed in the previous section, liaison is only employed, or licensed, in
certain linguistic environments. These environments are organized into two categories,
Obligatory, or those situations in which liaison is required, and Optional, or those
situations in which liaison licensing depends largely on speaker style and/or speech

situation. An example Of obligatory liaison has already been mentioned. In man ami
[m5nami] ‘my friend’, liaison is Obligatory. Liaison is always licensed after deterrniners.

In an optional liaison, the Speaker may choose whether to employ the liaison, and as such,
sociolinguistic considerations may apply. As will be discussed later, cases that constitute
Optional liaison are highly debated between researchers. Impossible liaisons can be

defined as those situations in which liaison never occurs. For example, liaison never

occurs after a Singular noun, as in Jean en a [366m] ‘John has some’.

The occurrence or nonoccurrence of liaison seems to involve numerous factors
from a diversity of areas in linguistics. These areas include phonology, phonetics,
prosody, syntax, and semantics. Each of these disciplines has attempted to describe
liaison from within its own unique framework. In the subsequent paragraphs there will
be a brief introduction to liaison within the frameworks of the various aforementioned
disciplines in addition to a discussion of research whose focus is the description of
appropriate liaison usage in speech, as well as research relating to the teaching of liaison,

learner errors in employing liaison, and implicit/explicit. learning. For the purposes of the

current study on liaison teaching and production, the following discussion will be mostly
limited to answering the question of when liaison occurs, not of how it occurs or why.
2.1.1 Phonology of Liaison

In terms of phonology, liaison causes the pronunciation of a latent consonant, as

in the aforementioned example in which [n] in man ami [m'énami] is realized
phonetically as Opposed to man chat [méja] where it is not. For liaison to occur, the

linking consonant must precede a vowel-initial or glide-initial word (T ranel 1987: 171).1
According to Valdman’s phonological account Of liaison in 1976 (99-105), liaison
consonants are latent in their underlying form. They become phonetically realized
before vowel-initial words and are deleted before consonant-initial words. For example,
this account would argue that the underlying form of the first person plural pronoun,

‘nous’ is InuZJ‘, with the liaison consonant being realized before vowel-initial words as

in ‘nous avons’ [nuzav5] ‘we have’ and remaining latent before consonant-initial words

as in ‘nous chantons’ [nuj‘dt'o] ‘we sing’. Other researchers that retain this view Of

liaison as an account of consonant-deletion before consonant-initial words include
Schane (1968) and Dell (1973). This view contrasts with other research studies
completed by Klausenburger (1974, 1984) and Tranel (1981), who attribute liaison
occurrence to the insertion Of a consonant before vowel-initial words, rather than the
deletion Of one before consonant-initial words.

In addition to consonant insertion and deletion accounts of liaison, non-linear
phonological accounts such those completed Clements, et al. (1983) and also by Tranel
(1995) use the prosodic structure Of words in order to understand liaison. Non-linear

accounts will be examined in more detail in the discussion on prosody.

4

2.1.2 Phonetics and Orthography
Tranel (1987:174-178) gives a relatively thorough account for the phonetic

occunence of liaison with each Of the liaison consonants. In his account, liaison

consonants are limited to the set, [t,z,n,rr,p,g]. Some accounts use [k] instead of [9], but
Tranel contends that [k] is archaic (19872174). Ayres-Bennett et al. (2001 :61) also
include [v] as a liaison consonant, citing as proof examples such as neuf ans [ntevd]

‘nine years’ and neufheures [ntevrerr] ‘nine o’clock’. However this seems to be a case

of enchainement and assimilation rather than liaison. Enchainement occurs when a word-

final consonant attaches to the empty onset of the following vowel-initial word. The

syllable boundary for neuf ans is represented by a period [nrevd]. Liaison involves
encha’inement, but [I] is not a liaison consonant since it is also phonetically realized
before consonants as in neuf chats [mafia] ‘nine cats’. In this case [f] is a non-latent
coda as opposed to the latent [z] in des chats [deja] ‘sorne cats’. The latent [z] in des
[de] ‘some’ cannot attach to the onset of the following word chats Ua] ‘cats’ since the
onset position is filled with [j] as in [deJa]. In neuf chats the non-latent [f] is attached to
the coda position of neuf as in [nrefJa]. In neuf ans the [f] consonant is still non-latent

and resyllabifies to the onset position of ans. Additionally, the [f] consonant undergoes

assimilation, a process by which a phonetic sound takes on characteristics, such as

voicing, of the immediately surrounding phonetic sounds. In the example of neuf ans, the

[-voiccd] [t] consonant takes on the characteristic of [+voiced] from the following vowel
[d] and becomes realized phonetically as [v], thus giving the final pronunciation [nrevd].

As a rule spelling indicates the phonetic nature of the linking consonant. Table 2.1

illustrates the distribution of phonetic liaison consonants according to their orthographic

representation. Ifliaison is to occur, orthographic <t> and <d> will produce [t], <S>,
<x>, and <z> will produce [2], <n> will produce [11], <r> will produce [K], <p> will

produce [p], and <g> will produce [9], or possibly [k] in some accounts.

Table 2.1 Orthographic/Phonetic Liaison Consonants
Orthographic Phonetic Examples
Representation Representation
t [t] petit enfant

[Pfitfifd]

‘small child’

d [t] grand enfant

[grfitfifdl

‘tall child’

S [2] des oranges

[dezcrrr‘i3]

‘ some oranges’

x [2] am: Etats- Unis

[ozetazyni]

‘in the Unite States’

2 [z] chez ewr

Lleza]

‘at their house’

n [n] ban anniversaire

[bananiverrsr-zrr]3

‘happy birthday’

r [g] premier autamne

[pxemjerroton]

‘first autumn’

p [p] trap heureux

[meme]

‘too happy’

g [g] or {1;} an long été
[élégete]

‘a long summer’

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

2.1.3 Liaison Consonants and Grammatical Categories

Grammatical context can also be indicative of which liaison consonant is to be
employed. Table 2.2 lists grammatical contexts for which [2] is employed as the liaison
consonant. When liaison occurs, [2] will .be the phonetic consonant employed for plural

determiners, adjectives, and nouns, singular prenominal adjectives and personal pronouns

ending in <s>, and the first and second person singular and plural conjugated verbs.

Perhaps due to its widespread occurrence in the French language, the liaison consonant

[z] is used widely in false liaisons, such as vingt-cinq années ‘twenty-five years’

[vétsékzane] (Tranel, 1987: 171). In this case of false liaison, the [2] functions as a

 

 

 

 

 

 

 

 

 

plural marker.
Table 2.2 Grammatical Contexts Utilizing [z]
Grammatical Context Examples
Plural determiners des arm's
[dezami]
‘some friends’
Plural adjectives petits enfants
[Pfizfifd]
‘little children’
Singular prenominal adjectives ending an gros éléphant
in orthographic <s> [égrozelefd]
‘a fat elephant’
Plural noun + adjective les enfants aimables
[lezr‘if'azemabl]
‘the pleasant children’
Personal pronouns nous, Mons, r'ls, elles, Nous avons du temps.
’95 [nuzavm]
‘We have some time.’
First person verbs Je suis arnérr'cain.
Bosqizomerrik‘é]
‘I am American.’
Second person verbs Vous chantez encore
[vuj‘dtezfikorr]

 

‘You are still singing.’

 

Table 2.3 shows the grammatical contexts in which [t] is the preferred liaison

consonant. This liaison consonant is phonetically realized in third person verbs, adverbs

ending in ‘—ment,’ and also in certain prenominal, singular adjectives such as petit

‘small’ and grand ‘tall’. This consonant has also been known to occur in false liaison.

For example, the insertion of [t] between a verb and pronoun, as in a-t-il [atil] ‘does he

 

have’, is a false liaison from the 16" century that is now commonly accepted as an
appropriate liaison (T ranel 1987:170).

Table 2.3 Grammatical Contexts Utilizing [t]

 

 

 

 

 

Grammatical Context Examples
Third person verbs Ils dannent un bonbon.
[ildcntéb’ébb]
‘They give a piece ofcandy.’
Adverbs ending in —rnent Extrémement important
[ekstrremomdt‘éporrtfi]
‘extremely irnportant’
Prenominal adjectives ending in un grand arbre
orthographic <t> or <d> [Eggfitagbg]
‘a tall tree’

 

 

 

The consonant [n] is also a frequently occurring liaison consonant. Table 2.4

shows the major contexts for which [n] is employed as the liaison consonant. These

include singular possessive determiners, singular indefinite determiners, and prenominal

adjectives, pronouns, prepositions, as well as monosyllabic adverbs that end in

orthographic <n>. Liaison with [It] only occurs in words that end in a nasal vowel.
However, a word that ends in a nasal vowel does not necessarily indicate that liaison will
occur with [n]. For example, in dans un mots [ddzémwa] ‘in a month’ the word dons
ends in a nasal vowel, but occurs in liaison with [z] (Tranel 19872175). In this case, the

liaison consonant produced is determined by orthographic <s>.

Table 2.4 Grammatical Contexts Utilizing [n]

 

 

 

 

 

 

 

 

Grammatical Context Examples
Singular possessive determiners ton qopareil—photo

[tfinaparrsjfoto]

‘your camera’
Indefinite Singular determiner un artiste

[‘énarrtist]

. ‘an artist’
Prenominal adjectives ending in ancien arnt'
orthographic <n> [Eisjenami]
. ‘former friend’

Pronouns ending in orthographic <n> On y va.

[‘5niva]

‘Let’s go.’
Prepositions ending in orthographic <n> en hiver

[finivarr]

‘in winter’
Monosyllabic adverbs ending in bien aimé
ormosraphic <n> [bj‘éneme]

‘well liked’

 

 

 

The remaining liaison consonants [3, p, g/k] are limited in their usage in the
spoken language. Table 2.5 lists grammatical contexts in which each of these consonants
is phonetically realized. The linking consonant [K] occurs with adjectives and is possible
after certain infinitives. The liaison consonant [p] is only used with trap ‘too much’ and

beaucoup ‘a lot’. The liaison consonant [g] (or [k] in some accounts) is only employed

in prenominal adjectives that end in <g>.

IO

Table 2.5 Rare Liaison Consonants

 

 

 

 

 

 

 

 

 

Phonetic Consonant Grammatical Context Examples
[3] Adjectives ending in an premier étage
orthographic (I) [oprromjerreta3]
‘on the second floor’
Certain infinitives aimer un peu
[mm]
‘to love a little’
[p] Adverbs trap and beaucoup trap heureux
[30mm]
‘too happy’
[9] or [k] Prenominal adjectives an long été
ending in <g> [‘él5g/kete]
‘a long summer’
2.1.4 Prosody

In regards to prosody, the liaison consonant can resyllabify with the following

vowel, guarding the preferred Consonant-Vowel (CV) syllable structure in French.

Recall the example using man amt ‘my friend’. The phonetic representation of this

phrase is [mfinami]. [ami] has an Open onset position since it begins with a vowel. In

liaison, the consonant [n] fills this position, leaving [m5] as an open syllable and thus

restructuring the syllables of [ami] to be [nami] in order to create the final structure

[m5.na.mi]. This allows for the retention of the preferred CV syllable structure for

French. This phenomenon is exemplified in non-linear phonological accounts since these

accounts utilize the prosodic structure on language to in order to comprehend liaison

OCCUTTCIICC.

In non-linear phonological accounts of liaison, the liaison consonant is treated as

a floating, or latent consonant (Ayres-Bennett 2001: 69). The liaison consonant is

floating in the sense that it has no skeletal tier with which to attach. Example (1) shows

ll

 

such a consonant. In des ‘some’, [(1] attaches to the onset slot and [e] to the rime slot, but
Since there is no Skeletal slot for [z] to attach to, it remains a floating, or latent,
consonant. For [2] to be realized phonetically, it needs to attach to a skeletal slot. In (2),
enfants [dfd] ‘children’ has an open onset slot. Consequently, when these two words

occur in succession, [z] attaches to the open onset slot of [dfd] to form the structure

represented in (3) and thus achieves phonetic realization.

(1) o

 

 

 

 

 

O R
l l
x x
| |
d e z
(2) o o
/\ /\
O R O R
| | l I
x x x x
l I |
(i f d
(3) o ' O o
O/\R 0/\R O/\R
I! x x x x x
| | | l |
d e 2 ii f EL

12

2.1.5 Syntax of Liaison

In addition to being understood in phonetics, phonology, and prosody, liaison can
also be understood in terms of syntax. This is reflected in the following example, which
is translated as ‘His friends have a house.’

(4) S

DPl/\VP
Der/\NPI V’/\ DP2
l | I /\

Ses N’ ont‘ Det NP2
\ | \\\ I I
amis \une N’
maison

In (4), liaison is shown between ses ‘his’ and amis ‘friends’ by a line connecting
the two words. Tranel reports that the stronger the syntactic tie, or syntactic solidarity,
that exists between two words, the greater the tendency for liaison to occur (1987:184).
There are two indications of the strong syntactic solidarity that permits liaison between
ses and amis. First, both ses and amis belong to the same syntactic constituent, DPl.
Syntactic solidarity is stronger within a constituent than between constituents. Thus,
liaison is more likely between ses and amis than between amis and ant. Secondly, the
notion of c-command provides firrther evidence of syntactic solidarity. C-command can
be defined as follows (Hawkins 2001 :307):

A c-commands B if and only if A does not dominate B and every category
dominating A dominates B.

Dominance can be defined as follows (Jones 199628):

Node A dominates Node B if there is a downward path from node A to node B
along the branches of the tree.

13

A path can be defined as “an unbroken series of branches and nodes that does not change
direction with respect to the top of the tree” (Culicover 1982:21).

Following these definitions, the determiner (Det) ses c-commands the noun
(N ’) amis because it does not dominate amis, yet every category that dominates ses,
namely DPl, also dominates amis. Thus, this furthers the idea of the existence of
syntactic solidarity between the two words. In the case of amis ‘friends’ and ant
‘have’, the noun (N ’) amis does not dominate the verb (V ’) ant, fulfilling the
dominance requirement for c-command, but not every category that dominates amis
dominates ant since DPl dominates amis but not ant. Thus, there is no c-command
relationship between amis and ant and liaison is not as likely to occur between these
words as it is to exist between ses and amis.

Prunet takes this examination further by using c-command to understand the
difference between obligatory and optional liaison. He states the following (Prunet,
1986:226):

If A and B c-command each other then liaison is obligatory.

If A c-commands B but the converse is not true, then liaison is Optional.
Using these parameters, liaison between ses and amis in (4) would be obligatory
because the two words c-command each other. Liaison between ant ‘have’ and one
‘a’ would be optional because the verb (V ’) ant c-commands the determiner (Det)
une but une does not c—command ant. This optional liaison is indicated by the

dotted line connecting ant and une.

l4

In terms of semantics, liaison can clarify the meaning of a phrase. Compare (5)
and (6), both adapted from Selkirk (1974: 59)“. Again, a line drawn between the two
words represents liaison.

(5) DP

un N PP
I
marchand
P NP
| I

de N’

N /\AP
| I
draps A’
anglais
(6) DP

P NP I
I I anglais
de T9 .
draps

These two phrases contain the same words, but the phrases express different meanings.
The representation found in (5) conveys the meaning, ‘a merchant Of English Sheets’ and

the representation in (6) conveys the meaning, ‘an English merchant of sheets.’ When

l5

liaison is produced between drops and anglais, it signals syntactic solidarity between the
two words and the unique representation found in (5). When no liaison is used, the
meaning is potentially ambiguous (Selkirk, 1974: 583,) and thus has the possibility Of
either the syntactic representation and consequent semantic interpretation given by (5) or
the one given by (6). Following is another example of liaison being used to clarify
semantic meaning, this time taken from Battye et al. (2000: 111):
(7) Elle donne I un bonbon.

[eldonéb5b5] 5

‘She gives a piece Of candy.’ (Third Person Singular)
(8) Elles donnent un bonbon.

[eldontrfib5b5]
‘They give a piece Of candy.’ (Third Person Plural)

In (8) the [t] is pronounced before the vowel-initial word un [tit] ‘a’, signaling that the

phrase is in the third person plural form, as opposed to the third person singular form
found in (7).
2.1.6 Contexts of Liaison

From the presentation above, it is evident that liaison has been examined by many
different areas of linguistics. Another important aspect of liaison research is liaison
licensing. Liaison licensing determines when liaison does or does not occur in Speech
and is a highly debated issue among researchers. Authors do agree, at least, on one item
of importance. They all agree that there exist certain linguistic contexts in which liaison
is obligatorily used, some in which it is never used, and some situations where its usage is
optional.

The following three tables (2.6, 2.7, 2.8) summarize what six researchers deem

obligatory, optional, and forbidden licensing of liaison. The accounts of Delattre

I6

(1966:43), Valdman (1976:43), and Tranel (1987:185-190) are considered as they
provide some “traditional” accounts of liaison. The accounts of Dumenil (2003: 16) and
Battye et al. (2000:110-111) are included as they form the textbook part of the cuniculum
used to teach liaison at Michigan State University, the base for the current study. Finally,
the results from a corpus analysis completed by Booij and de Jong (1987110) are being
presented as they contain an analysis of French liaison usage in Tours that is based on
actual speech and that diverges from all other accounts presented for consideration.
Examples are provided for each context and liaison is exhibited by the balding and
underiining of the consonants and vowels involved. The nonoccurrence of liaison is
exhibited by two forward slashes ‘//’ representing the blocking of liaison in impossible
cases.

Table 2.6 lists contexts in which these researchers have determined liaison is
Obligatory. A number has been assigned to each context in order to aid the reader in
locating the appropriate context for discussion. All researchers presented agree that
obligatory licensing occurs in the following sequences: determiner-noun (context 1),
deterrniner-adj ective (5), personal pronoun-verb" (8), verb-personal pronoun (10), and in
set phrases (15). However, there are many areas where these researchers diverge in their
accounts of obligatory liaison. These include whether numerals (2 and 3), the sequences
determiner-pronoun (4), adj ective-noun (6), adj ective-adjective (7), and pronoun-pronoun
(9) are obligatorily involved in liaison, as well as whether all pronouns or uniquely
personal pronouns may occur in liaison with verbs (8 and 10), and whether a ’est ‘it is’, il
est ‘it is’ (11 and 12), and monosyllabic adverbs (13) and prepositions (14) may occur in

liaison at all.

17

In comparing these accounts, it appears that Dumenil aligns most closely with
Battye et al. This is important to note because these two accounts form the textbook
input for the FRN 330 French Phonetics and F RN French 430 French Linguistics courses
respectively. It is also relevant to note that although Dumenil] most closely aligns with
Battye et al., there are still differences between these two accounts. For example, Battye
et al. do not discuss numerals (2 and 3) and they limit the number of monosyllabic
adverbs (13) and prepositions (14) that occur in obligatory liaison. Additionally,
Dumenil does not discuss determiner-proform sequences (4). The differences in these
two accounts could be potentially confusing to these learners as they take the two courses
using these textbooks.

Where researchers disagree may be due to a variety of reasons. First,
disagreements may be due to the way in which the syntactic categories were labeled and
organized. For example, Valdman’s account holds that liaison occurs between adjectives
(7), as in de vraimciens documents, or ‘genuine old documents’. While Delattre does
not seem to agree with Valdman’s claim because he does not include the adjective-
adjective sequence in his list of obligatory contexts, he does include de vraiwciens
amis, or ‘real old fiiends’ in the category of “determiner + adjective.” Thus, this seems
to be a representational difference and not a true disagreement regarding when liaison is
employed. Additionally, this adjective-adjective sequence is a rare occurrence as the
number of prenominal adjectives is low and their possibility for occurring in liaison is
even lower. Consequently, other researchers may not have considered it imperative to
include it in their descriptions of obligatory liaison. Secondly, dissent among researchers

may be due to differences in types of corpora analyzed. Booij and de Jong are the only

18

researchers listed in Table 2.6 that give information about the corpus they analyzed. No
other researcher presented provides background information on the corpora they used for
their analyses, nor do they give any indication that they used a corpus for analysis.
Finally, the differences in these accounts may be due to a genuine disagreement about
when liaison is obligatorily licensed. Whatever the case may be, it is important that
researchers find an agreement on liaison licensing in order that it be correctly presented

to second language learners.

19

Table 2.6 Obligatory Liaison Licensing

 

Context
number

Delattre
(1966:43—48)

Valdman
(1976: 106)

Tranel
(1987: 185-190)

Battye et al.
(2000:110-112)

Dumenil
(20032164, 170,172)

Booij and
de Jong
(1987:1010)

Examples

 

Det. + Noun

Det. + Noun

Det. + Noun

Det. + Noun

Det. + Noun

Det. + Noun

degnfants
“some children’

 

Numeral + Noun

un ‘One’, deux ‘two’,
trais ‘three’, six
‘six’, neuf‘nine’, dix
‘ten’ +_

deufinfants
‘two children’

 

numeral + et um “and
one’+ ans ‘years’ or
heures ‘hours’

trenle et uflns
‘thirty-one years Old’

 

Det. + Pron.

Det. + Pron.

Det. + Proform

 

 

Det. + Adj.

Det. + Adj.

Det. + Adj.

Det. + Pron.

leflulres
‘the others’

 

Det. + Adj.

 

Adj. + Noun

Adj. + Noun

Det. + Adj.

Det. + Adj.

un ancien ami
‘an old friend’

 

Adj. + Noun

 

Adj. + Adj.

Adj. + Noun

un petit_enflmt
‘a small child’

 

 

 

 

Pers. Pron. + Verb

Pron. + Verb

Pron. + Verb

de vraifinciens documents
‘genuine old documents’

 

Proform + Verb

 

Pers. Pron. +
Pers. Pron.

Pron. + Pron.

Pron. + Pron.

Pron. + Verb

Pers. Pron.
+ Verb

vaus_étes
‘you are’

 

Pron. + Pron.

 

Verb + Pers. Pron.

Verb + Pron.

Verb + Pron.

 

c ’est (impersonal) + _
‘it is’ +

c’est +NP
‘it is’ + NP

c ’est (impersonal) + d
‘it is’ +_

 

il est (impersonal) + _
‘it is’ +_

il est (impersonal) + NP

‘it is’ + NP

il est (impersonal) + _
‘it is’ +_

 

Monosyl. Adv. + _

Monosyl. Adv. + NP

trés + _
‘Very’ + _

 

Monosyl. Prep. + m

Monosyl. Prep. + NP

dons ‘in’,

. C ' 7
sans Without,
sous ‘under’,
chez ‘at/to’ ,
en ‘in’ +_

 

 

Set Phrases

 

 

Set Phrases

 

Set Phrases

 

 

Verb er

‘Very’ + _

dons ‘in’,

sans ‘without’,
sous ‘under’,
chez ‘ai/to’ ,

en ‘in’ i;
”SHIV

 

Pron. + Pron.

vaugy étes
‘you’re here’

 

Verb + Pron.

allefl
‘go on’

 

 

LL..\

6 ’est_rm avacat
‘he’s a lawyer’

 

iI étaiggnefais
‘once upon a time’

 

Monosyl. Adv. + _

trés_rgtile
‘Very useful’

 

Monosyl. Prep. +_

danfln livre
‘in a book’

 

Set Phrases

 

Set Phrases

 

Etatggnis
‘United States’

 

 

 

Optional liaison is highly dependent on speaker style and speech Situation.
People tend to use liaison less frequently when involved in an informal conversation and
more frequently when speaking formally. The correct usage of Optional liaisons can
Si gnal mastery of the French system (Battye et al. 2000: 112) Table 2.7 presents differing
accounts of Optional liaison. Dumenil does not give a syntactic account of optional
liaison and thus will not be included in the current discussion.

Notice the vast divergence between the accounts presented in Table 2.7.
Complete agreement is not met on any syntactic context. Delattre, Tranel, Battye et al.,
and Booij and de Jong agree that liaison is optional after plural nouns (1) and after verbs
(4). Delattre, Tranel, and Battye et al. agree that liaison is optional after polysyllabic
adverbs (7), and polysyllabic prepositions (8). However, the extent to which researchers
agree ends here. Even the “traditional” accounts of Delattre and Valdman completely
disagree. The vast disagreement found in the case of Optional liaison may be due its
variable nature since it is largely dependent on speaker style and speech Situation.
However, it is important to reiterate the point that this disagreement may also be due to
differences in corpora used for analysis. The need for comprehensive research in liaison

becomes even more evident with impossible liaison.

21

Table 2.7 Optional Liaison Licensing

 

Context
Number

Delattre
(1966:43—48)

Valdman
(1976: 107)

Tranel
(1987: 185-190)

 

Battye et al. ,4 Booij and de Jong
(2000:110,112) (1987:1011)

 

1

P1. Noun + _

Pl. Noun + _

Examples

 

Pl. Noun +_ Inflected head +
Complement

 

 

Pron. (not personal) + _

des personnes dge'es
‘some older people’

 

d ’autreflrriverant
‘others will arrive’

 

Post—posed Pers. Pron. + _

avans-naufln livre
‘do we have a book’

 

Verb + _

Verb +
Complement

Verb + Inflected head +
Complement: Complement

je vaisgssayer
‘I’m going to try’

 

_ + Past Participle

ils n ’ant rieflcheté
‘they bought nothing’

 

Pl. Adj. + invariant word

bangu mavais
‘good or bad’

 

Polysyl. Adv. + _

Adv. +
Complement7

Polysyl.
Adv. +_

sauvent_absenl
‘often absent’

 

Polysyl. Prep. + _

Preps. +
Complement8

Prep. +_

pendant_une semaine
‘during one week’

 

pas, trap +_

not’ ,
‘too much’ +_

trap_l_reureux
‘tOO happy’

 

Monosyl. Conjunctions + _

quand
(subord. conj.) + _

mais +_
‘but’ +_

maifllars ‘but then’
quaniil arrivera “when he arrives’

 

quand, comment +
est-ce que,

 

quand, comment in
indirect questions

Quanflst-ce an ’1‘] arrivera?
‘When will he arrive?’

 

Savez-vaus camment_ils reviendrant?
‘DO you know how they will come back?’

 

Present forms of
étre ‘to be’

 

-ant third person
plural ending

je suiflméricain
‘I’m American’

 

 

ilsfanLgn effort
‘they’re making an effort’

 

Set Phrases

 

a brafluverts
‘to open arms’

 

Complement +
Complement

 

 

 

 

 

 

des attaques nucléaireflme’ricaines
‘American nuclear attacks’

 

Inflected Word +
Non—complement

 

 

 

ils re’fle’chissent_avant de répandre
‘they think before answering’

 

 

 

 

The domain of impossible liaison is exhibited in Table 2.8. Recall that impossible
liaisons are marked with two forward Slashes ‘//’ to Show that liaison is not possible
between the two words being examined. These are contexts in which liaison is never
licensed. Research presented by Booij and de Jong, as well as that presented by
Valdman, do not give explicit accounts of forbidden liaison, although it might be
assumed that impossible liaison would constitute any context not covered by their
presentations of obligatory and Optional liaison. The authors are not included in Table
1.8 due to their lack of overt discussion of impossible liaison contexts. Delattre,
Dumenil, Battye et al., and Tranel agree that liaison cannot occur after a singular noun
(1), after the conjunction at ‘and’ (8), and before words that begin with an h-aspiré (12).
Delattre, Tranel, and Dumenil agree that liaison is impossible after interrogative adverbs
such as quand ‘when’ and comment ‘how’ (9)”. Delattre, Battye et al., and Dumenil
agree that liaison is also impossible in certain fixed, or set phrases (11). Dumenil mostly
aligns with Delattre, although differences exist in their treatments of pronouns (4, 5, 6, 7),
polysyllabic conjunctions (10), numerals (13), glides (14), and the affirmative response
word, aui ‘yes’ (16). T ranel’s account provides the smallest number of cases for
impossible liaison, but this may be due to the fact that he largely concentrated on when
liaison actually occurs in language. He provides a list of Obligatory liaison contexts, but
only mentions contexts of impossible liaison when needed for clarification. The accounts
Of Dumenil and Battye et al. give two different descriptions of liaison usage, which is
important to note because again these two presentations form the textbook input for the
French Phonetics 330 and French Linguistics 430 courses respectively. The two accounts

do not necessarily disagree per se, but they do give parameters for liaison usage that are

not consistent with each other. For example, Battye et al. do not treat proper nouns (2),
complex plural nouns (3), post-posed personal pronouns (5), or interrogative adverbs (9)
while Dumenil’s account does not treat sense groups or pauses ( 15) in terms of liaison
and only lists liaison with out ‘yeS’ as forbidden in the case of mais oui ‘but yes’ (16) and

says nothing tO the effect of liaison being impossible in cases Of emphasis.

24

Table 2.8 Impossible Liaison Licensing

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Context Delattre Tranel Dumenil Battye et al. Examples

number (1966:43-48) (1987: 185-190) (2003:165-168,172) (2000:110-111)

I Sing. Noun + _ Sing. Noun + _ Sing. Noun + w Sing. Noun + _ un saldat // espagnal
‘a Spanish soldier’

2 Proper Noun + _ PrOper Noun + _ Jean // a

_ ‘John has’

3 Complex Pl. Nouns Complex Pl. Nouns des salles // a manger
‘dining rooms’

4 Pron. (not pers.) ending in a Chacun // y va
nasal vowel + _ ‘each one is going’

5 Post-posed Pers. Pron. ending Post—posed Pers. Pron. + _ qu ’a-t-an // afaire
in a nasal vowel + A _ ‘what is one to do’

6 Post-posed 3rd Pers. Pron. + _ ant-ils // un livre

_» ‘dO they have a book”

7 on, ils, elles + Past Part. 01) sont-elles // allées

#_ ‘where did they go’

8 at + _ at +_ et +_ 6! +_ an hamme et // unefemme’
‘and’ ‘and’ ‘and’ — ‘and’ ‘a man and a woman’

9 Interrogative Adv. quand ‘When, quand ‘when’, comment quand // est-i] venu

comment ‘how’ ‘comment’ (except in ‘when did he come’
in direct questions Comment allez vaus? ‘How
are you?),
cambien ‘how much’ +
inversion #—
10 Polysyl. Conj unctions + _ néanmains // elle resta
‘nevertheless she stayed’
11 Set Phrases Set Phrases Set Phrases unefais // au 1 ’autre
‘One time or another’
12 _ + h-aspiré words _ + h-aspiré words _ + h-aspiré words _ + h—aspiré un // héras
_ words ‘a hero’

13 _ + un ‘one’, huz't ‘eight, anze _ et un + _10 ‘_ and one, as 1m ‘one’, huit dans // anzejaurs
‘eleven’, uniéme ‘first’, in thirty one’, _ + anze ’Cight’, anze ‘in eleven days’
huitieme ‘eighth’, anzieme ‘eleven’ ‘eleven’

‘eleventh’ .
14 _ + (some) glides ‘ + (some) glides [es //yuccas
_ ‘the yuccas’
15 Across a sense ils se sant leve’s // c‘z cinq heures
group boundary ‘they got up at five o’clock’
_ 0r pause

16 + aui mais oui When emphasis i] dil // aui

:+ ‘yes’ ‘but yes’ or Clarity is used ‘He said yes’

 

 

 

 

 

25

 

 

 

 

 

 

As evidenced by the discussion above, liaison context is a debated topic within
French linguistics. The implications this has for teaching are important in that in order to
promote acquisition of liaison, accurate and native-like input needs to be provided for
learners. Table 2.9 presents a summary of the individual contexts of obligatory and

impossible liaison in which researchers agree. These will form the contexts to be used in

 

 

 

 

 

 

 

the present study.
Table 2.9 Obligatory and Impossible Liaison Contexts
Obligatory Examples Impgssible Examples
determiner + noun deunfants singular noun + _ an soldat // espagnol
‘some children’ ‘a Spanish soldier’
determiner + uwcien ami et + _ un homme et // unefemme
adjective ‘an old friend’ ‘and’ + __ ‘a man and a woman’
personal pronoun + vouLétes h—aspiré words an // he’ros
verb ‘you are’ ‘a hero’
verb + pronoun alien set phrases une fois // on I ’autre
‘go on’ ‘one time or another’
set phrases EtatL-flnis
’United States’

 

 

 

 

Due to the variable nature of optional liaison and the fact that there are minimal accounts
of agreement among researchers as to which contexts of liaison are optional, this type of
liaison will not be treated in the current study. Additionally, optional liaison is not
treated by Dumenil in terms of syntactic contexts. Since her account forms the basis of
input for participants in the study, it would not be useful to test participants on optional
liaison for the purposes of the current study.
2.2 Teaching Liaison

Thus far, the discussion has centered on understanding the phenomenon of liaison
through various areas of linguistics, as well as understanding when liaison occurs.
Another area of concern with regards to liaison is the teaching methods used to teach it.

There appears to be very little research done in this area. In 1973, Olsen proposed and

26

 

implemented a plan for a corrective phonetics course in which she contended that
teaching must focus on oral usage of liaison and not just explicit liaison rules. In her
discussion of liaison, she argued that instructors should focus on oral practice in addition
to the explicit teaching of rules. However, she provided no description of what types of
oral practice activities should be used nor did she provide empirical analysis of the effects
her teaching methodology had on liaison production by L2 learners. In Tranel’s account
of liaison (1987:168-190), he suggested that second language learners should only learn
obligatory contexts. However, here again there was no information provided on methods
that should be used to teach liaison contexts nor did he present information examining the
effects of teaching only obligatory contexts on liaison production by L2 learners. In
1998, Thomas completed a study of a French immersion course in Canada. He proposed
an alternative to liaison instruction that typically involves complex textbook
presentations. He suggested that instructors adopt the following two rules for teaching
liaison: (I) teach liaison only with consonants [z], [t], and [n], and (2) only teach
obligatory liaisons, including fixed expressions. His basis for these rules is that liaison
usage is continually declining and consequently only the most common liaisons should be
taught Again, there was no information offered on teaching methods that should be used
nor empirical research testing the effectiveness of following his two rules for liaison
teaching. In order to truly know the effectiveness of a teaching method, empirical
research should be done. This type of research has largely been ignored in regards to
liaison.

In addition to considering research relating to instruction, it is important to

consider errors committed by L2 learners employing liaison in speech. Errors can prove

27

to be a useful tool to teachers because they can highlight concepts that need clarification.
Additionally, information about the types of errors committed can provide insight into the
second language acquisition process for liaison.

There have been numerous studies that look at error correction, but very few that
have recorded the types of errors actually committed by learners. In 1953, Gaudin
published a study that accounts for liaison errors in American learners of French. In her
report, which is now over 50 years old, she gives a partial list of examples of liaison
errors made by these learners (19532460). This list corresponds to the following syntactic
situations: between personal pronouns and verbs, after monosyllabic prepositions and
adverbs, between prenominal adjectives and the nouns they modify, and in set phrases.
She claims that learners produce liaison in forbidden cases such as after et ‘and’, before
h-haspiré words, and after singular nouns. However, she provides no empirical support
for her analysis. Additionally, she gives no account for methods used to determine
errors. According to Walz (1980:424), errors have typically been treated perceptually
and not empirically in research. These treatments, in which the researcher relies on
his/her personal impressions of what errors are commonly being made by learners, may
in fact ignore error patterns that could provide useful information to language teachers. If
errors are improperly determined because they are based on perceptions, then it is
possible that some errors may go unnoticed and fossilize in the L21eamers'
interlanguage. Tranel (1987: 188) contends that learners tend to err in making too many
liaisons rather than making too few. He states that English speakers learning French have
a tendency to produce more liaisons than needed due to pronunciation conventions in

their native language. Because most written letters in English are pronounced, they

pronounce most written letters in French. Again, there is no empirical support provided
for this analysis. In his 1998 study, Thomas gave some indication of empirical findings,
but still fell short of taking into account individual occurrences of learner errors. Recall
that Thomas studied a French immersion course in Canada. He observed that learners did
not use liaison in 20% of obligatory cases (547-548). This is a more helpful examination
of learner errors than a perceptual study, but it still does not give any information the
specific types of errors committed.
2.3 Explicit and Implicit Instruction

Finally, as this study will be looking at the effectiveness of explicit instruction,
there needs to be some discussion of implicit versus explicit learning. Implicit learning
can be defined as the “acquisition of knowledge about the underlying structure of a
complex stimulus environment by a process that takes place naturally, simply and
without conscious operations” (Ellis 1994:1). The learner usually cannot explain the
knowledge rules acquired through this type of learning experience. For example, a
learner might hear a new structure, enter it into memory, use it accurately in a language
situation, but not be able to explain why or how the structure works. Ellis defines explicit
learning as “a more conscious operation where the individual makes and tests hypotheses
in a search for structure” (Ellis 1994: l). Explicit knowledge is attained when a learner
searches for and builds a repertoire of information on a structure and then tests
hypotheses or when a learner assimilates a rule after explicit instruction (Ellis 1994: 1-2).
In explicit instruction, focus is put on learner input such that the learner receives direct
explanation and practice of a language feature. For example, the group receiving explicit

instruction in the current study will be told directly what the rules are that govern liaison

29

production and then they will practice identifying appropriate grammatical situations for
liaison to occur as well as situations where it should not occur.

While knowledge is attainable through both implicit and explicit learning, there is
much debate as to which cognitive abilities are acquired implicitly and which are learned
explicitly (Ellis l994:2). According to Gass and Selinker (2001:209), simple and salient
features are most easily acquired implicitly. However, a study completed by Green and
Hecht (1992: 179) supports the idea that explicit instruction is most effective for learning
simple language features such as word order. Additionally, Gass and Selinker (2001:209)
note that the effects of explicit instruction on complex language features are unclear.
Liaison, with its high variability of usage among native speakers, would fall towards the
complex side of the simple-complex language feature continuum. The present study will
examine the effectiveness of explicit'instruction on a language feature that falls into this
area.

In 2002, Hulstijn argued that “explicit learning and practice often form efficient
ways of mastering an L2 by creating opportunities for implicit learning” (193). In a study
completed in 1997, MacWhinney contended that both implicit and explicit processes
contribute to language learning. Researchers seem to agree that both implicit and explicit
processes are important in learning a language. The current study will look to see what
effect implicit input and/or explicit instruction has on the liaison production of L2
learners.

2.4 Conclusion
In summary, liaison has been examined in numerous linguistic disciplines, namely

phonology, phonetics, prosody, syntax, and semantics. Liaison usage falls into three

30

categories. These categories are obligatory, optional, and impossible. There is little
agreement among researchers as to what types of syntactic contexts fall into each of these
categories. For the ease of the reader, the contexts in which all of the researchers reached
agreement are reprinted here in Table 2.10.

Table 2.10 Researcher Agreement

 

 

 

 

 

 

 

 

Obligatory Impossible
determiner + noun singular noun + _
determiner + adjective et + _

‘and’ + _
personal pronoun + verb h-aspire words
verb + pronoun set phrases
set phrases

 

 

More research needs to be done in this area in native French liaison usage in order that
L2 learners be provided with accurate input in their courses. In terms of teaching liaison
there is little known about the effects of teaching methods on liaison production or on the
types of errors committed by L2 learners. Research in both these areas is necessary in
order to find effective ways to teach liaison to L2 learners. Implicit and explicit learning
are important processes to language acquisition and as such, continued research in this
area will benefit both second language acquisition research and instructors as they search
for better ways to meet the needs of their students.

Liaison is an integral part of the French language. Ifnative-like fluency is the
goal of language study, then appropriate language teaching methods are needed in order
to provide learners with the necessary resources to approach the level of fluency. For
language learners who wish to be proficient in French, acquisition of liaison is key, for
proper employment of it can indicate a mastery of the language system (Battye et al.

20002112). This desire for language proficiency, in addition to the lack of empirical

31

research examining the effects of teaching methods used to teach liaison on liaison
production by second language learners, have become the driving force for the current
study. The explicit teaching method, currently in use at MSU, will be examined in order
to study its effects on liaison production by L2 learners. In addition to information on the
effectiveness of the explicit teaching method, this study will hopefully provide insight on
how learners use liaison in their speech in order to highlight areas that may need more

emphasis in teaching.

32

CHAPTER THREE

EXPERIMENTAL DESIGN

3.1 Research Questions and Hypotheses

Recall that the research questions for the present study are as follows:

1.

2.

How does an explicit teaching method influence liaison production in second
language speech?

How do second language learners of French employ liaison in speech?

In order to answer these questions, the following seven hypotheses have been formed.

1.

There will be a significant difference between groups (Implicit, Instruction,
Implicit + Instruction) in terms of appropriate, or correct liaison production.

There will not be a significant-difference between liaison production and speech
style (Paragraph, List).

. There will be a significant difference between the number of obligatory and

impossible liaisons produced by learners.

There will be no significant difference between the number of liaisons produced
in determiner-noun and determiner-adj ective sequences.

There will be a significant difference between the number of liaisons produced in
personal pronoun-verb and verb-pronoun sequences.

There will be a significant difference between the number of liaisons produced in
obligatory set phrases and impossible set phrases.

There will be a significant difference between the number of liaisons produced
with the consonants [z], [t], and [n] as the liaison consonant.

The first hypothesis serves to answer the first research question relating to explicit

instruction and liaison production since it test differences between instruction treatments.

A significant difference is expected to occur between the groups in this case since groups

receiving instruction should produce liaison in a more native-like manner than the group

receiving no instruction.

33

The last six hypotheses relate to the second research question regarding how
students employ liaison in speech since they look at how learners use liaison in various
instances. No significant difference is expected to occur between liaison production and
speech environment since students should produce the same number of obligatory and
impossible liaisons irrespective of reading style. Rules governing obligatory and
impossible liaison do not change based on reading style. Thus, learners are expected to
produce approximately the same number of liaisons in the contextualized reading section
as they do in the decontextualized phrase list.

In regards to obligatory and impossible liaison usage, a significant difference is
expected because learners should be exposed to liaison occurring in obligatory contexts
and should receive little to no examples of impossible liaison in input. Thus, there should
be more occurrences of liaison in obligatory contexts than in impossible contexts in L2
speech.

Determiner-noun and determiner-adjective sequences should show no significant
difference because determiners are quite common in the French language and always
occur in liaison”. Students would have frequently encountered speech in the course of
their French study that included determiners occurring in liaison.

A significant difference is expected between personal pronoun-verb and verb-
pronoun sequences even though they both occur obligatorily in liaison. This is expected
for two reasons. First, personal pronoun-verb sequences seem to occur more frequently
in the French language than verb-pronoun sequences. Secondly, personal pronoun-verb

sequences are taught earlier than verb-pronoun sequences. Thus students should have

34

been exposed to enriched input including personal pronoun-verb sequences occurring in
liaison.

The difference between the number of liaison occurrences in obligatory and
impossible set phrases is expected to be significant since thesepphrases should occur in
input categorically with liaison in respect to obligatory cases and categorically without
liaison in respect to impossible cases. Learners are expected to produce liaison in
obligatory set phrases and to refrain from producing liaison in impossible set phrases
because should have little to no evidence in their input to the contrary.

Finally, the liaison consonant [z] is expected to be the most frequently employed
liaison consonant since it is the most frequently occurring liaison consonant in French

speech (Tranel 1987: 176). Due to the pervasiveness of [z], these learners probably have

been exposed most frequently to the occurrence of this liaison consonant in their input.
3.2 Method

3.2.1 Experiment Variables
The independent variables are instructional treatment (implicit, instruction, implicit +
instruction), reading style (paragraph, phrase list), and time (pretest, posttest). The
dependent variable is correct liaison production.

3.2.2 Participants12

Participants were selected from a volunteer base gathered at Michigan State

University. This volunteer base was formed using four courses: two FRN 321 Oral
Expression courses, one FRN 340 Survey of French Literature course, and one F RN 330
French Phonetics course. Volunteers were selected to participate if they did not have any

of the following characteristics: participation in an immersion program, participation in a

35

study abroad program, visitation to a French speaking country or province for longer than
six weeks, or previous participation in FRN 330 French Phonetics or an equivalent
course.

The 21 respondents (males = 4, females = 17) were all native speakers of English
between the ages of 18 and 21 (mean age = 19.5). All reported to be studying French
either for personal enrichment, for future employment opportunities, or both. They spent
an average of 20 minutes speaking in French and an average of 45 minutes practicing
pronunciation weekly outside of class. Only three people participated in extracurricular
activities involving the use of French.

Participants were divided into three experimental groups based on their
educational background information. Each group was labeled according to type of
instruction received.
3.2.2.1 Implicit Group (IMP)

This group consisted of eight students (males = 1, females = 7) that received only
implicit input. These participants were taking either FRN 340 Survey of French
Literature or FRN 321 Oral Expression. In these courses, discussion of L2 readings was
a primary focus and these learners did not receive any formal teaching of liaison but were
exposed to it in implicit input presented in the speech of their instructor, who was a
native speaker of French.
3.2.2.2 Instructional Group (INS)

These eight participants were taking FRN 330 French Phonetics and received
explicit instruction13 of liaison. Liaison was mentioned in passing as examples occurred

in other lessons14 and then formally taught with rules and practice during four class

36

periods of 50 minutes each. The instructor, who was a native speaker of French, spent
four 50-minute periods on explicit instruction on liaison, one 50-minute period on review,
and then one 50-minute period on written testing. Approximately 31% of class time was
spent on written transcription exercises in which students identified both obligatory and
forbidden liaison situations, 28% on oral pronunciation activities in which students read
phrases aloud that included both obligatory and impossible liaison situations, and the
remaining 41% was spent on lecture. Both oral and written exercises were assigned as
homework. The textbook used for liaison instruction was Facile a dire! (Dumenil, 2003).
3.2.2.3 Implicit + Instructional Group (IMPINS)

The IMPINS group represented participants that received both formal instruction
and extra implicit input. These participants were taking FRN 330 Phonetics in addition
to FRN 340 Survey of French Literature or FRN 321 Oral Expression. These students
received the same formal instruction as the INS group in addition to receiving implicit
input from either the FRN 340 or the FRN 321 courses.

The pretest served as a way to equate the three groups. The groups showed a
normal distribution and a one-way ANOVA on the total correct score indicated that there
was no significant difference between the groups. Thus, the groups were at comparable
levels with regards to liaison usage prior to instruction.

3.2.3 Raters

Two raters analyzed the data for the presence of liaison. One rater was a native
speaker of French and the second rater was a native speaker of English. They were
provided with written instructions and were given opportunities to ask questions two days

before examining the data.15 They were instructed to listen for the presence of a liaison

37

consonant in the targeted environments and were allowed to listen to the data as many
times as necessary before recording their observations.
3.2.4 Test Medium"

Since it is difficult to elicit needed data from learners in a conversational style
setting, the researcher constructed a test composed of two parts, a reading passage
(paragraph) and a phrase list (list). Liaison situations were examined within sentential
context in the paragraph section and without sentential context in the list in order to
examine whether obligatory and impossible liaison production could be related to speech
style. The paragraph section consisted of a letter written by a girl to her aunt. The letter
was constructed by the researcher and then approved by a native speaker of French. The
test was organized such that all participants read first the reading passage and second the
phrase list. The test components were organized in this way for practicality reasons. In
’. the first task, participants were asked to read the paragraph section for understanding and
afterwards they were given an opportunity to ask any comprehension questions. This was
done to familiarize the participant with the passage in order to aid in fluency when
reading on tape. It seemed more pragmatic to complete this task using the paragraph
reading section since it provided a framework in which participants could negotiate the
meanings of vocabulary items. The same test was used for the pretest and posttest so as
to reduce the threat to validity due to test equality. The pretest and posttest should be
comparable since they are the same test. The syntactic contexts chosen for testing were
taken from the agreement among researchers in regards to liaison licensing discussed in

the previous section. These contexts are repeated in Table 3. l.

38

Table 3.1 Obligflory and Impossible Liaison Contexts

 

 

 

 

 

 

 

 

 

Obligatory Impossible
determiner + noun singular noun + _
determiner + adjective et + _

‘and’ + _
personal pronoun + verb h-aspiré words
verb + pronoun set phrases
set phrases

 

These contexts were chosen as they reflect the agreement between the researchers
presented in the last section and also the input that the INS and IMPINS groups were to
receive in their textbook. Words in the categories of ‘set phrases’ and ‘h-aspiré’ were
chosen directly from the textbook (Dumenil 2003: 164, 166-167) used in FRN 330
Phonetics so as to ensure that participants were being tested on information that was
present in their input. Because these words are specific cases of liaison usage or
nonoccurrence, as opposed to entire syntactic categories (i.e. determiner-noun), these
words need to be memorized. The data set used for obligatory contexts is found in Table
3.2 and the data set used for impossible contexts is found in Table 33. The specific
obligatory liaison that is being examined is underlined and the impossible liaison is
represented by two bars ‘ //’ placed in between the words in question. Each category
contains four phrases with the exception of the set phrases. The set phrases had to be
limited to two for obligatory contexts and two for impossible contexts because there were

only two set impossible phrases presented in the textbook input.

39

Table 3.2 Obligatory Context Data Set

 

Determiner + Noun

moaactivite’ préfere'e
‘my favorite activity’

 

lemrés-midis
‘ the afternoons’

 

Sowtivite’ préfere’e
‘ his favorite activity’

 

Cesgnimaux
‘ these animals’

 

Determiner + Adjective

uzuimmense jardin
‘an immense garden’

 

moweien jardin
‘my former garden’

 

deghorribles bétes
‘ some horrible bugs’

 

desenormes e’le’phants
Some horrible elephants

 

Personal Pronoun + Verb

iLLe’taient
‘ they were’

 

"OMOHS
‘ we have’

 

ellesgvaient
‘ they had’

 

ellemnt
‘ they have’

 

Verb + Pronoun

Viennentils chez nous?
‘ Are they coming to our house?’

 

Sonhils d l ’universite’?
‘Are they at the university?’

 

Que choississentils comme me’tier?
‘ What are they studying?’

 

travaillengjls toujours ensemble
‘ do they still work together’

 

Set Phrases"

Etats-_Unis
United States

 

 

 

petiLd petit
‘little by little’

 

 

Table 33 Impossible Context Data Set

 

Singular Noun + __ un lulin // ancien
‘an antique gnome’

 

Manon // a
‘Manon has’

 

Marlin // a
‘Martin has’

 

an enfant // insupportable
‘an unbearable child’

 

et + an lutin ancien et // artisanal
‘an antique handmade gnome’

 

Jean et // Anne
‘John and Anne’

 

Marlin, Manon, Sebastien et // Alice
‘Martin, Manon, Sebastian, and Alice’

 

et // elles m’ont dit
‘and she told me’

 

_ + h-aspiré‘s petit // hamac
‘little hamack’

 

les // hamburgers
‘the hamburgers’

 

les // haricots"
‘the beans’

 

des beaux // hibow:
‘some beautiful owls’

 

Set Phrases”0 a tort // et // a travers
‘Wildly’

 

 

nez Ila nez
‘face to face’

 

 

3.2.5 Questionnaire21
Upon completion of the posttest, participants belonging to the INS and IMPINS
groups were provided with an anonymous questionnaire asking for their perceptions on
how the various classroom activities helped them recognize and produce liaison
appropriately.
3.2.6 Procedure
Participants were scheduled individually to complete the pretest and posttest. The

tests took place approximately five weeks apart so as to reduce the threat to validity due

4]

 

to test recollection. The pretest took place approximately two weeks before instruction
and the posttest took place the week after instruction. During the initial appointment,
each participant was furnished with a consent form and a code to protect his/her identity.

Before taping, the researcher read through the task instructions with the
participant.22 The participant was asked to first read through the paragraph section
silently and was then subsequently provided an opportunity to ask questions about the
passage. This provided each participant with an opportunity to clarify unclear vocabulary
meanings and also allowed them to familiarize themselves with the reading passage in
order to aid in fluency when being recorded on tape. Next they read the paragraph
section and list section on tape. All taping took place in a language lab equipped with
tape recorders and headsets at individual stations.

On the pretest, there was an odd occurrence where a participant produced an
atypical consonant in liaison. This participant produced an [s] as opposed to a [z] in the
phrase elles out [8125] ‘she has’. This is simply a difference in voicing and so was treated
as a legitimate attempt at liaison production. Another participant produced a [d] as
opposed to [z] in the phrase Etots-Unis [etazyni] ‘United States’. This was also treated
as a legitimate attempt at producing liaison because this is simply a difference in
airstrearn features. [d] is [-continuant, -strident], while [2] is [+continuant, + strident].
This phenomenon also occurred once in the posttest. One participant produced
[potidamak] as opposed to [pctiamak]. Albeit an incorrect usage of liaison, this was

treated as a legitimate attempt at producing it because this difference is simply in voicing.

42

 

After the data was recorded, the primary researcher averaged the scores of the
raters and then analyzed them statistically using Minitab. Interrater reliability was
calculated using the Pearson Product-Moment Correlation. A correlation of 0.89 existed
between the two raters. This is a significant correlation (p<.001). This is appropriate
considering the rating task relied on the raters’ perception of a pronounced consonant.

A three-way ANOVA test was conducted in order to look for the existence of
significant improvement in the three groups with regard to overall correct liaison
production, as well as to see if there was a significant difference in speech style. Tukey
post hoc tests were used to pinpoint specific differences between variables. The posttest
was used for testing hypotheses dealing with how L2 learners produce liaison in speech
acts since it represented the most recent sampling of learner data. One-way ANOVAs
were calculated in order to compare obligatory and impossible liaison usage, determiner-
noun and determiner adjective sequences, personal pronoun-verb and verb-pronoun
sequences, obligatory set phrases and impossible set phrases, and also the liaison

consonants [z], [t], and [n]. A General Linear Model AN OVA was used for each test

because it is a robust test that can handle unbalanced designs.

43

CHAPTER FOUR
RESULTS AND DISCUSSION

Caution must be taken when interpreting these results. The small group size
limits generalizeability, so the conclusions drawn from the results may be true for these
learners uniquely. Each hypothesis will be restated for the convenience of the reader and
discussed individually.
4.1 Hypothesis One: Teaching Methods

The first hypothesis states that there will be a significant difference found
between groups (IMP, INS, IMPINS) in terms of liaison production. In order to test this
hypothesis, the number of appropriate, or correct, liaison productions and the number of
appropriate refrains from production were calculated and then totaled in order to give an
overall correct liaison production. Table 4.1 compares the number of correct liaisons
produced during the pretest with the number produced during the posttest.

Table 4.1 Correct Liaison Production versus Time

 

 

 

 

 

Number of Correct Liaison Productions
Group Implicit (IMP) Instruction (INS) Implicit + Instruction (IMPIN S)
Pretest 348/512 342/512 235.5/320
68.0% . 66.8% 73.4%
Posttest 357/512 380/512 246.5/320
69.7% 74.2% 76.9%

 

 

 

 

 

It is important to note that the IMPINS group only had 320 possibilities to produce
liaison because that group contained three less people than the other groups. Table 4.1
shows that there is very little change in score between the pretest and posttest for any of
the groups. A three-way ANOVA revealed that there was, in fact, no significant
difference between the pretest and posttest for the groups and no significant interaction

between group (IMP, INS, IMPINS), time (pretest, posttest), and speech style (reading,

list). Additionally, there was no significant difference found between the three groups at
the posttest. These results suggest that the explicit teaching method employed had no
effect on liaison production for these students. However, remember that these results
need to be treated with caution due to the small number of participants.

A closer examination of the data using a Tukey post hoc test reveals that there
appears to be a tendency for the IMPINS (Implicit + Instruction) group to produce liaison
more correctly than the IMP (Implicit) group (T=2.2802, p=.065). This shows that
learners receiving implicit input in addition to explicit instruction tended to produce
liaison more correctly than those only receiving implicit input.

Given these results, it is possible that the combination of implicit input and
explicit instruction was more helpful for these learners in developing better or more
native-like patterns of liaison usage than merely receiving implicit input was able to
accomplish. More research will need to be done in order to appropriately validate this
claim for a larger group of students.

4.2 Hypothesis Two: Speech Style

The second hypothesis states that there will be no significant difference between
number of liaison occurrences and speech style (Paragraph, List). Table 4.2 compares the
number of correct liaison productions for each speech style.

Table 4.2 Correct Liaison Production According to Speech Style

 

 

 

 

 

Number of Correct Liaisons
Speech Style Implicit Instruction Implicit + Instruction
Paragraph 171/256 176.5/256 118.5/ 160
66.8% 68.9% 74.1%
List 186/256 203.5/256 128/ 160
72.7% 79.5% 80.0%

 

 

 

 

 

45

A two-way ANOVA and Tukey post hoc test on correct liaison usage versus
reading style and group indicated that there was a significant difference between the
paragraph reading and list (F =13.88, p<.01) in all three groups and that no significant
interaction occurred between reading style and group. Overall, there was a higher
number of more appropriate, or correct liaisons produced in the phrase list than in the
paragraph reading. This may suggest that the liaison production of these learners is
dependent on speech style. This could be explained by Labov’s description of the effects
of different reading styles on pronunciation. He states that “people have little conscious
control over their use of variables in reading style. The actual content of the test is more
influential” (1972:81). It appears that people pay more attention to understanding content
when reading a passage in paragraph form. Word lists, because they direct the
participants' attention to one word at a time, may produce more careful speech than a
reading passage (Trudgill, 2000286). Thus, these learners may have paid more attention
to their pronunciation during the word list activity, consequently producing a higher
number of correct liaisons and more attention to content during the reading activity,
resulting in a lower number of correct liaison productions.

4.3 Hypothesis Three: Obligatory versus Impossible Liaisons

The third hypothesis states that there will be a significant difference between the
number of obligatory and impossible liaisons produced by learners. Table 4.3 compares
the total observed liaison production, the correct liaison production, and incorrect liaison

production for obligatory and impossible contexts.

Table 43 Obligatory and InHJossible Production

 

 

 

 

 

Number of Obligatory Number of Impossible
Productions Productions
Observed 498/756 96/588
65.9% 16.3%
Correct 498/756 492/588
65.9% 83.7%
Incorrect 258/756 96/588
34.1% 16.3%

 

 

 

A one-way ANOVA revealed a significant difference between the number of

obligatory and impossible liaisons produced by these learners (F =51.97, p<.001). Of the

756 opportunities to produce an obligatory liaison, these learners produced a total of 498

of them, or 65.9 percent. Of the 588 potential impossible liaisons that could be produced,

learners produced a total of 96, or 16.3 percent. These learners appear to produce liaison

most frequently in situations that are obligatory, but still do not produce it in all the

situations in which it should be produced. Additionally, they appear to make

overgeneralization errors by incorrectly producing liaison in impossible contexts.

However, these learners do appear to be making fewer errors in impossible contexts than

in obligatory contexts. They incorrectly produced liaison in impossible contexts 16.3

percent of the time, while they incorrectly chose not to produce liaison in obligatory

contexts 34.1 percent of the time. From this data, it appears that these learners are

making fewer production errors in impossible contexts than in obligatory contexts. This

is inconsistent with Tranel’s claim that learners with an L1 of English tend to err in over-

producing liaison as opposed to under-producing it (19872188). It is possible that these

learners have focused more on impossible liaison situations than on obligatory situations

and consequently produced fewer errors in impossible contexts than in obligatory

contexts.

47

 

4.4 Hypothesis Four: Sequences with Determiners
The fourth hypothesis states that there will be no significant difference between
the number of liaisons occurring in determiner-noun and determiner-adj ective sequences.

Table 4.4 shows the number of liaisons produced in determiner-noun and determiner-

 

 

 

 

 

adjective sequences.
Table 4.4 Deterrniner-Noun versus Determiner-Adjective
[Grammatical Context iaison Licensing bserved Usage
' 124.5/168
eterrniner-Noun bligatory 74. 1%
104.5/ 168
lDeterrniner-Adj ective IObligatory 2.2%

 

 

 

Of the 168 opportunities for liaison to be produced in a determiner-noun
sequence, learners actually produced liaison 74.1 percent of the time while they produced
liaison 62.2 percent of the time in determiner-adj ective sequences. A one-way ANOVA
revealed that there is in fact, no significant difference between liaisons produced in these
two sequences. This seems appropriate because learners should have been exposed to
input that includes the consistent occurrence of determiners occurring in liaison since
they fall into the obligatory category of liaison. Additionally, determiners are among the
first syntactic categories learned in French. Since liaison after determiners is
obligatory”, learners should have been exposed to very little input to the contrary.

4.5 Hypothesis Five: Sequences with Verbs and Pronouns

The fifth hypothesis states that there will be a significant difference between the
number of liaisons produced in personal pronoun-verb sequences and verb-pronoun
sequences. Table 4.5 shows liaison production in pronoun-verb and verb-pronoun

sequences.

Table 4.5 Personal Pronoun-Verb versus Verb-Pronoun Sequences

 

 

 

 

 

matical Context iaison Licensing bserved Usage
125.5/168
Personal Pronoun-Verb IObligatory 74.7%
85/168
Verb-Pronoun [Obligatory 50.6%

 

 

 

A one-way ANOVA revealed that there is a significant difference between
personal pronoun-verb sequences and verb-pronoun sequences produced by these
learners (F =-3. 63, p<.05). These learners produced significantly more liaisons in
personal pronoun-verb sequences than in verb-pronoun sequences. They produced
liaison in personal pronoun-verb sequences 74.7 percent of the time while they only
produced liaisons 50.6 percent of the time in verb-pronoun sequences. This difference is
possibly due to the fact that personal pronoun-verb sequences are generally taught earlier
and they seem to occur more frequently in speech than verb-pronoun sequences.

4.6 Hypothesis Six: Set Phrases

The sixth hypothesis states that there will be a significant difference between the
number of liaisons produced in obligatory set phrases and impossible set phrases. Table
4.6 shows liaison productions for obligatory and impossible set phrases.

Table 4.6 Obligatory versus Impossible Set Phrases
IGrammatical Context [Observed Usage

 

 

Lin/84
IObligatory Set Phrases 9.6%
l10.5/s4

possible Set Phrases 12.5%

 

 

 

A one-way ANOVA revealed that there is a significant difference between the
number of liaison occurrences in obligatory set phrases and the number of liaison
occurrences in impossible set phrases produced by these learners (F = 4.30, p<.01).

These learners produce significantly more liaisons in obligatory set phrases than

49

impossible set phrases. These learners produced liaisons 69.6 percent of the time in
obligatory set phrases, while only producing them 12.5 percent of the time in impossible
set phrases. Again, this is possibly due to input presented to students. Liaison in
obligatory set phrases should be categorically present in learner input and liaison
occurrences in impossible set phrases should be categorically absent from input.
Additionally, these learners seem to be making fewer production errors in
impossible set phrases. They made production errors only 12.5 percent of the time in
impossible set phrases and 30.4 percent of the time in obligatory set phrases. These
results are consistent with those obtained for the hypothesis discussed in section 4.3,
comparing obligatory and impossible contexts as a whole. Consequently, the same
explanation that follows the results in 4.3 should also apply here, namely that these
results are inconsistent with Tranel’s claim that learners with an L] of English tend to err
in over-producing liaison as opposed to under-producing it (1987:188), and that learners
could have focused more on impossible liaison situations than on obligatory situations
and consequently produced fewer errors in impossible contexts than in obligatory
contexts.
4.7 Hypothesis Seven: Liaison Consonants

The seventh hypothesis states that there will be a significant difference between
the number of liaisons produced with [z], [t], and [n] as the liaison consonant. Table 4.7
shows the correct, observed liaison production of these learners according to the liaison

consonants [z], [t], and [n].

Table 4.7 Liaison Production According to Liaison Consonant

 

 

 

 

Liaison Consonant Observed Usage IIndividual/Total Usage
276/378 276/484.5

| 2] 73.0% 57.0%
103.5/210 103.5/484.5

”t1 49.3% 21.3%
105/168 105/484.5

[n] 62.5% 21.7%

 

 

 

 

A one-way ANOVA revealed no significant difference between the number of
liaisons correctly produced with the consonants [z], [t] and [11]. Out of 378 obligatory
contexts in which the last consonant of the word was <z>, <x>, or <s>, these learners
produced a [z] consonant a total of 276 times, or 73.0 percent. Out of 210 contexts in

which the last consonant of the word was <t> or <d>, learners produced [t] a total of

103.5 times, or 49.3 percent. Out of 168 contexts in which words ended in <n>, learners
produced [11] a total of 105 times, or 62.5 percent. The [2] consonant comprised 57.0
percent of all liaisons produced. The [t] consonant comprised 21.3 percent of all liaisons

produced. The [n] consonant comprised 21.7 percent of the total liaisons produced.

These results are consistent with Tranel’s observation that [z] is the most frequently

occurring consonant in liaison (1987:176). However, it is important to note that this

difference may be due to the fact that half of the 756 words that were tested ended in <z>,

<x>, or <s>, the orthographic letters that encourage the production of [2], thus giving

learners more opportunity to produce it than to produce [t] or [n]. Words ending in <t>

or <d> comprised 27.8 percent of the total words and words ending in <n> comprised

only 22.2 percent.

SI

 

4.8 Further Discussion

At the posttest respondents who participated in the INS and IMPINS groups were
asked to fill out an anonymous questionnaire“ regarding their perceptions of how
classroom activities and instruction have helped them recognize and produce liaison.
Participants rated a series of ten statements on a likert scale ranging from one to five,
with one being “Strongly Disagree” and five being “Strongly Agree.” There were five
statements relating to liaison recognition and five statements relating to liaison
production. Eight statements concerned the four activities completed during liaison
instruction, namely the textbook presentation, lecture, written transcription, and oral
reading. Two statements concerned other French courses being currently taken and
consequently these statements were only responded to by the IMPIN S group. The
breakdown of participant responses is shown in Tables 4.8 and 4.9. Table 4.8 shows how
students rated the effectiveness of each of the activities in aiding them to recognize and
produce liaison appropriately. Table 4.9 reports ratings on how helpful the various
activities were to these learners in producing liaison in appropriate, or correct, situations.

Table 4.8 Recognition of Liaison

 

 

 

 

 

 

 

 

 

 

 

Textbook Lecture Written Oral Other
Presentation Transcriptions Reading Courses”
Strongly Agree 6 5 4 4 0
Agree 6 7.5 6 7 3
Neutral 0 .5 2 2 2
Disagree 0 0 1 0 O
Strongly Disagree 1 O 0 0 0

 

 

 

52

Table 4.9 Production of Liaison

 

 

 

 

 

 

 

Textbook Lecture Written Oral Other
Presentation Transcriptions Reading Courses“
Strongly Agree 3 2 l 7 0
Agree 6 9 6 4 3
Neutral 4 2 4 1 2
Disagree 0 0 1 1 O
Strongly Disagree 0 0 1 0 0

 

 

 

 

 

 

 

These students most strongly agreed that the textbook presentation was most
beneficial in helping them to recognize liaison situations and that reading sentences aloud
was most beneficial in helping them to produce liaison correctly. In regards to overall
agreement, these students perceived that both the textbook presentation (12 total) and the
lecture (12.5 total) were helpful to them in perceiving when liaison should or should not
be produced and most students indeed found every activity helpful to some degree.

There were only two participants that felt that an activity (textbook presentation and
written transcription) were not beneficial in helping them to recognize when liaison
should occur.

Additionally, these students most strongly agreed that the oral reading exercise
was the most beneficial activity for helping them to produce liaison in appropriate
situations. In regards to overall agreement, these students perceived that both the oral
reading exercise and the lecture were helpful to them in producing liaison in appropriate
situations. There were three students that felt that activities (written transcriptions,
reading aloud) were not beneficial in helping them produce liaison.

It is important to note that while these students perceived that these activities were
helpful, the results of the posttest indicate that these activities did not significantly
improve students’ production of‘liaison in appropriate situations. This disconnect

between the insignificant improvement in liaison production and these leamers’

53

perceptions maybe due to the fact that these activities provided the first overt teaching of
liaison that these learners experienced and as such, they may have perceived the activities
to be more helpful than they actually were.

Each question asking for the participant’s perception of how effective the activity
was in helping them to recognize appropriate liaison situations had a corresponding
question relating to its effectiveness in aiding correct liaison production in speech. A
Pearson Product-Moment correlation was nm on each pair of questions in order to see if
these learners felt that the activity had the same effect on their ability to recognize and
produce liaison. There was a significant correlation found between two sets of questions.
The first set corresponds to questions three and eight on the questionnaire.

3. In the French 330 Phonetics course, the written transcription exercises helped me
recognize when liaison should occur.

8. In the French 330 Phonetics course, the written transcription exercises helped me
to correctly produce liaison in speaking situations.

A significant correlation was discovered between the responses of these two questions
(r = .61, p<.05). A breakdown of the responses to this question is presented in Table
4.10.

'ons 3 8
"on Production

Table 4.10

       

S

Neutral

S
Overall, these learners perceived that the transcription exercises were equally effective in

helping them recognize and produce liaison. The other significant correlation occurred

between questions four and nine (1' = .81, p<.001). These questions are repeated below
and the breakdown of responses is shown in Table 4.11.

4. In the French 330 Phonetics course, reading sentences aloud in class helped me
recognize when liaison should occur.

9. In the French 330 Phonetics course, reading sentences aloud in class helped me to
correctly produce liaison in speaking situations.

4 9
' 'on Production

Table 4.11

     

S

Neutral

Overall these learners perceived that reading sentences aloud helped them equally
recognize and produce liaison.

Surveys such as the one discussed can be helpful to teachers because they give
information about what activities students feel are most productive in helping them learn.
However, care must be taken in interpreting surveys such as these. The survey should be
compared with student output in order to examine the relationship between their
perceptions and their ability to employ the structure correctly. Although these learners
perceived that certain activities aided them in liaison recognition and/or production, there
was no significant improvement in their actual production on the test. It may be the case
that since this was their first overt experience with liaison, they felt that certain activities

were more helpful than they actually were.

55

CHAPTER FIVE
CONCLUSION
Recall again the research questions for the current study:

1. How does an explicit teaching method influence liaison production in second
language speech?

2. How do second language learners (L2) of French employ liaison in speech?

While it is not possible to draw general conclusions about the data presented due
to the small group sizes, the data do bring some interesting insights into the acquisition of
liaison. It seems that the results indicate that the explicit teaching method does not
significantly improve the correct liaison usage of these learners. The results of the
IMPINS group show a tendency for these learners to use liaison more correctly than the
IMP group. It is possible that if the experiment is repeated with larger group sizes then
the explicit teaching method may be found to be statistically significant in improving
liaison usage among learners when used in combination with extra implicit input. It is
also possible that a simple increase in input may produce the same results.

The second research question has been treated by comparing learner usage of
liaison in obligatory versus impossible contexts, determiner-noun sequences versus
determiner-adj ective, personal pronoun-verb sequences versus verb-pronoun sequences,
obligatory set phrases versus impossible set phrases, as well as in terms of liaison
consonant choice. These learners produced significantly more liaisons in obligatory
settings than in impossible settings, which may suggest that they are in the process of
acquiring liaison. Because they produced liaison in impossible, or forbidden, situations
as well, they may have been making overgeneralization errors. However, it is important

to note that these Ieamers were actually better at appropriately restraining from

employing liaison in impossible contexts than producing liaison in Obligatory contexts.
This may indicate that they placed more focus on learning impossible liaisons than
obligatory liaisons. Learners also produced significantly more liaisons in personal
pronoun-verb sequences than in verb-pronoun sequences. Because these are both
obligatory cases of liaison, this may suggest that more emphasis needs to be put on verb-
pronoun sequences in teaching liaison. Learners produced significantly more liaison in
obligatory set phrases than in impossible set phrases, which further supports the idea that

learners are producing significantly more liaisons in obligatory contexts overall than in

impossible contexts. Finally, these learners produced liaisons most ofien with the [z]

consonant.

The anonymous questionnaire revealed that these learners strongly agreed that the
textbook presentation helped them the most in recognizing liaison. They also strongly
agreed that the oral reading activity helped them the most in producing liaison. They
perceived the transcription exercises to help them equally in recognizing and producing
liaison. Additionally, they perceived the oral reading activity to also help them equally in
recognizing and producing liaison. Although these learners perceived these activities to
be helpful, there was no significant improvement in their actual liaison usage in speech.
This may be due to the fact that these activities provided the first overt teaching of liaison
that these learners experienced and as such, they may have perceived the activities to be
more helpful than they actually were.

There are a few evident limitations on this study. The group size limits the
generalizeablility of results. Any conclusions drawn from this data may only apply to

these learners. Also, the reading style is a potential confound because the paragraph and

57

list styles produced statistically different results. Finally, the mode of analysis by the
raters may be a limitation because it relied on raters’ perceptions of the presence of
liaison consonants between words. When phrases are pronounced quickly, it can be
difficult to perceive whether a liaison consonant was produced.

As for additional research, it is necessary to replicate this study with more
participants and more raters in order to achieve more accurate and generalizeable results.
Also, a replication study might consider using acoustic analysis to determine the presence
of a liaison consonant. Doing this will avoid issues of rater perception. It would also be
interesting to test another teaching method in regards to its effect on liaison production.
This would prove helpful to teachers as they search for better methods to meet their
students’ needs. Additionally, a longitudinal study that looks at the effects of teaching
liaison contexts incrementally, as part of their introductory French curricula, might prove
to be another useful study for teachers. The effects of studying abroad on liaison
production would provide insight into the importance of enriched input on interlanguage
development and liaison production. There also needs to be more research completed on
the process of the acquisition of liaison by L2 learners. Finally, it would also be useful to
complete a time-series design study on learners in order to gain a more complete picture

of the process of liaison acquisition.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

ON ON 252 2NN .ONN 222a N N42 N2 22 O N22223:
ON ON 2.22.2 222222 2NN .ONN 2222 2 N 2 ON .2 N 222222
NV ON 252 ON .ONN 2222 N N2 2 N2 2 N N22223:
OO ON 262 O2.N ON 222 N 2 22 O2 2 N 22.22222
ON2 Nw 282 ON .ONN 2222 2 2 2 ON .2 2 N22:
OO OO 252 ONN 222.2 2 2 0 ON 2 O2 N22
OO N2 232% ONN 222.2 O N.2 ON 2 N N22
O2-N ON 252 ONN 2.222 N 2 :2 N 2 O N22
NN2-OO O 252 ONN 2222 N N2 0 O2 2 N N22
ON ON 252 ONN 2222 2 2 ON 2 2. N22
OO 2 252 ONN 2.22.2 O 2 N2 2 N N22
OO O 252 ONN 222.2 2 2 ON 2 N N22
O2: O 252 ONN 222 O 2 ON ”2 2 N22
8.2 O 282 2NN 2222 O N 0 2 a O .222
ON ONOO 282 2NN 2222 O 2 O2 22 N .2222
N Nv 262 2NN 222.2 2 2 O2 ..2 N 2222
2 8oz 2NN 2.22.2 2 Nw2 6 ON 22 O .222
O ON 252 ON 222 2 2 0 ON "2 N .222
2 ON 252 2NN 22222 O 2 0 ON 2 N .2222
O2 ON 22222 55:2 2NN 222.2 N N2 2 ”2 N .222
O2. ON 282 O3. 2222 N 2 o 2N 22 2 2222
xoogowscg xoogogcmz m0£>fl0< mow—300
222 328822 222 328822 22022022 22822260 2264.00». 22222on 02660
22222222222285 wee—822m 3226.522622be 222922220 222625222 a 22622223262 new: 0922 xom 222222222322
82222222822 22222222352 2.< O2222;
20332220222 2.222222032222222

< EDZmE:

59

APPENDIX B
EXPLICIT INPUT FROM TEXTBOOK (Dumenil 2003:161-180)
The chapter on liaison begins with a short history on word-final consonants
followed by a definition of liaison. Liaison is defined as follow (162):

Liaison is the linking of a final consonant that is not normally pronounced with
the initial vowel of the following word.

Two examples are given for comparison:

deuxfilles ‘two girls’ [dofij] versus deux amis ‘two friends’ [dezami]
un petit garcon ‘a small boy’ [Eptigarrso] versus an petit outil ‘a small tool’ [‘éptituti]

After defining liaison, Dumenil discusses how liaison can both clarify meaning
and cause ambiguity. Following are two of the examples she gives (162-163):

Example of clarification:
C ’est or)? ‘Where is it?’ [seu]
C ”est tout? ‘Is that all?’ [setu]

Example of ambiguity:
Elle a an petit ami. ‘ She has a boyfriend.’ [elaéptitami]
Elle a un petit tom's. ‘She has a small sieve.’ [elaéptitami]

Next, Dumenil presents some phonetic effects of liaison (163). Among these are
included denasalization, voicing and devoicing, and syllabification. In regards to
denasalization, she states that nasal vowels denasalize in liaison with the exception of
mon ‘my’, ton ‘yours’, son ‘his’, on ‘one’ (impersonal), en ‘in’, un ‘one/a’, bien ‘well’,
rien ‘nothing’, aucrm ‘any’, commrm ‘common’. She gives examples of both the general
rule of denasalization and the exceptions:

un bon éle‘ve ‘a good student’ [Ebonelev] (denasalization)
mon ami ‘my fiiend’ [monami] (exception

In regards to voicing and devoicing, Dumenil states that occlusives devoice and fricatives

become voiced in liaison, stating the following examples (163):

un grand ami ‘a tall friend’ [Egratami] (devoicing)
des amis ‘sorne friends’ [dezami] (voicing)

Finally, in terms of syllabification, Dumenil states that the liaison consonant re-
syllabifies to the following syllable in the majority of liaison cases. She gives the
following example (163):

les amis ‘the friends’ [le.za.mi]

Learners are encouraged to make a strong effort to re-syllabify the liaison consonant to
the following word.

Afier discussing the linguistic background on liaison, Dumenil presents cases of
obligatory liaison. She tells learners that liaison is only made between words belonging
to the same rhythmic group, but also that belonging to the same rhythmic group does not
necessarily indicate that liaison will occur between two words. She tells learners they

need to memorize the following cases where liaison is obligatory (164):

61

 

Table 3.1 Obligatory Liaison

 

Context Examples

 

determiner +noun un ami ‘a friend’, mon ami ‘my fiiend’, les amis ‘the friends’,
des amis ‘some friends’, ces ami ‘these friends’, deux amis ‘two
friends’, un petit ami ‘a boyfiiend’, quells amis ‘which friends’,
an ancien ami ‘a former friend’

 

pronoun (pronoun) + ils ont ‘they have ’, ils en out ‘they have some’, ont-ils ‘do they
verb/(pronoun) + verb have’, en ont ils ‘Do they have some?’

 

+ pronoun
fixed expressions les Etats- Unis ‘the United States’, les Champs-Elysées ‘the
(partial list) Champs-Elysées, petit a petit ‘little by little’, accent aigu ‘acute

accent’, avant hier ‘before yesterday’, sous entendu
‘insinuation’, commun accord ‘common agreement’, tout a
l’heure ‘in a little while’, Comment ollez-vous? ‘How are you?’,
quand est-cc que ‘when is it that. . . ’, tout c‘r coup ‘suddenly’, tout
dfait ‘quite’, de plus en plus ‘more and more, de temps en temps
‘from time to time, de fond en comble ‘from top to bottom’, de
haut en has ‘from top to bottom’

 

 

after monosyllabic tre‘s intéressant ‘very interesting’, bien evident ‘well evident’,
adverbs and dans un livre ‘in a book’, en an instant ‘in a moment’
prepositions

 

 

Afier presenting cases for obligatory liaison, Dumenil notes that obligatory
occurrence is largely observed for the first three cases, while the last case (monosyllabic
adverbs and prepositions) is subject to variation.

At the end of the section on obligatory liaison, there are five sentences in which
learners are asked to repeat, transcribe, and mark the obligatory liaisons. Following are
the first two of them (165):

1. [Is sont allés at Rome en avion. ‘They went to Rome by plane.’

2. II n ’est pas trés intelligent, mais il est amusant. ‘He is not very intelligent,
but he is fun.

After the transcription exercises, Dumenil presents cases of impossible liaison.
She begins by stating that liaison is forbidden between words belonging to different
rhythmic groups. She then goes on to state that within rhythmic groups there are some

classes of words for which liaison is forbidden. She begins by discussing h-aspiré. She

62

 

states that French has two types of <h>, one of which permits liaison, and one that does
not. She gives heure ‘hour’ as an example that permits liaison and héros ‘hero’ as an
example that does not. She then provides a partial list of h-aspiré words for learners to

memorize (166):

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Table B.2 H-aspire'

H-aspiré word Translation
la hache - the axe

la hachisch the hashi sh
la hate the hedge

la haine the hate

les Halles the Hallcs

le halo the halo

le hamac the hammock
le hameau the hamlet

le hamburger the hamburger
la hanche the haunch
le handball the handball
le handicap/les handicqpe's the disabled
le hangar the shed

le harce‘lement the harassment
Ie haricot the bean

la home the harp

le hasard the chance
la hausse the increase
la hauteur the height

la hernie the hernia

le héros the hero

le hibou the owl

la hiérarchie the hierarchy
le hippie the hippie

le hockey the hockey
la Hollande Holland

le homard the lobster
la honte the shame

le hoquet the hiccup

le huit the eighth

le hurlement the howling

 

 

After presenting the list of h-aspiré words, Dumenil makes note of some false liaisons

that have been known to occur in speech. Among them she lists:

63

 

Les haricots ‘the beans’ I"[lezariko]

Il va aller ‘He is going to go’ *[ilvazale]

Quatres amis ‘four friends’ *[katrrozami]

Il ira d toi ‘He will come to you’ *[rrratatwa]

Dumenil continues the discussion of impossible liaison by stating other cases for which

liaison is never licensed (167):

 

 

 

 

 

 

 

 

 

Table B.3 Impossible Liaison
Context Examples
Singular Noun + _ un soldat // anglais
‘an English soldier’
After et et // il
‘and he’
After a proper noun Jean // a dit
‘Jean said’
After a post-posed personal pronoun Ont-ils // un livre?
‘Do they have a book?’
After quand + inversion, after Quand // a-t-il dit qu ’iI partirait?
comment and combien ‘When did he sgy he will leave?’
Plural complex nouns and certain des salles // a manger
fixed expressions ‘dining rooms’
nez it nez
‘face to face’
Words beginning with <y> (except Les //yaourts
for les yeux ‘the eyes’), <w>, before ‘the yogurts’
the number eleven, and between the Les week-ends
conjunction mais ‘but’ and the ‘the weekends’

affirmative response word oui ‘yes’

 

Ils ont habité aux Etats-Unis pendant // onze ans.
‘They lived in the United States for 11 years.’
Mats // oui

‘but yes’

 

After presenting cases of impossible liaison, Dumenil provides five exercises for

practice. Learners are instructed to read aloud each phrase, transcribe it, and mark all

obligatory and impossible liaisons. The first two sentences from the exercises are

presented below (168):

 

l. Quand a—t-elle annoncé qu ’elle allait rendre les interrogations écrites?
‘When did she announce she was going to return the written questions?

2. Léon en a profité pour aller au magasin.
‘Leon made the most of his trip to the store.’

Dumenil continues her presentation of liaison by discussing cases where it is
optional. She informs learners that optional liaison is dependent on speaker style and
that as the formality of the speech situation increases, so does the number of optional
liaisons that occur in speech. She states that optional liaison fiequently occurs in political
discussions, in radio and television interviews, and in televised newscasts. She goes on
to say that the higher frequency of liaisons in the speech situations just mentioned is due
to the fact that they typically involve speech that has been practiced and is not

spontaneous (168). She then presents four cases for optional liaison (169):

 

 

 

 

 

 

Table B.4 Optional Liaison
Speech situation Type of liaison Example
occurrence
Theatrical performance All possible liaisons Deshommesillustresontottendu.
(poem reciting, a play, etc.) [dezomzilystrroz‘otatddy]
‘The illustrious men waited.’
Formal discourse Most liaisons De_s_'hommes_illustres ont_attendu.
[dezomzilysmotatfidy]
‘The illustrious men waited.’
Elevated conversation Some liaisons Dahommes illustres ont_attendu.
[dezomflystrr‘otatfidy]
‘The illustrious men waited.’
Familiar conversation Only obligatory De_s_hommes illustres ont attendu.
liaisons [dezamflystgfiatfidy]
‘The illustrious men waited.’

 

 

 

Dumenil then cautions learners to only make obligatory liaisons when participating in
informal conversation in order to achieve a more nafivelike pronunciation (169). She
ends the discussion on optional liaison by giving five exercises. She instructs learners to

repeat the phrases aloud, transcribe them, and mark the obligatory, optional, and

65

 

impossible liaisons. Following are the first two phrases she gives along with translations

(169-170):

1. Les Halles sont d Rungis depuis plusieurs années.
‘The Halles have been at Rungis for several years.’

2. Les autres étudiants que vous avez sont-ils aussi mouvais?
‘Are the other students that you have bad?’

Dumenil treats numbers as a separate entity with regards to liaison. She provides

some general guidelines for using numbers in liaison. She treats the numerals one, two,

three, six, nine, and ten in her account. Her presentation (170) is summarized in the

 

 

 

 

 

 

 

 

following table:

Table B.5 Numerals in Liaison

Numeral Occunence in liaison Example

un Obligatory after an but does not occur after J bi unenfant.

‘one’ vingt et un ‘twenty-one’, trente et un ‘thirty- ‘I have one child’
one’, etc. except if followed by ans ‘years’ or Ila vingt et un étudiants.
heures ‘hours’ ‘He has 21 children.’

deux Obligatory but some speakers do not make trente-deux étages

‘two’ liaison in cases of multiples of two such as ‘thirty-two floors’
trente-deux ‘thirty-two’ ,

trios Obligatory but some speakers do not make quarante-troisouvriers

‘three’ liaison in cases of multiples of three such as ‘forty-three factory
trente-trois ‘thirty-three’ workers’

six Obligatory but some speakers do not make vingt-siJLerreurs

‘six’ liaison in cases of multiples of six such as ‘twenty-six errors’
trente-six ‘thirty-six

neuf Obligatory before ans ‘years’ and heures neuLans

‘nine’ ‘hmr217‘s’, but occurs in enchainement everywhere ‘nine years’
else

dix Obligatory but some speakers do not make quatre-vingt digéglises

‘ten’ liaison in cases of multiples of dix such as ‘ninety churches’

 

quatre-vingt-dix ‘ninety’

 

 

Dumenil completes this section by providing an exercise consisting of ten

sentences in which learners are expected to repeat aloud and transcribe in writing,

 

marking the obligatory, optional, and impossible liaisons. The first two exercises are

given below (171):

1. Nous irons aux Blots-Unis le trente et un octobre d huit heures.
‘We will go to the United States on October 31 at 8 o’clock.

2. Jean et Marie vont essayer d ’aller a la plage avec elles.
‘John and Marie are going to try to go to the beach with them.’

Dumenil ends the chapter on liaison by providing a quick reference table for
obligatory and impossible liaisons and three sets of review exercises. The first exercise
requests learners to repeat twenty phrases, while inserting liaison where appropriate.
Following are the first two sentences of the exercise (173):

1. Les autres enfants arrivaient en autobus.
‘The other children were arriving by bus.’

2. Prends-en un outre, il y en a beaucoup.
‘Take another, there are a lot of them.’

In the second exercise, the phrases of the first exercises are given along with a phonetic
transcription. Students are requested to read each phrase according to its transcription.
Following are the first two sentences of the exercise (174):
1. Les autres enfants arrivaient en outobus.
[lezotrrozfifdarsivednotobys]
‘The other children were arriving by bus.’
2. Prends-en un autre, i] y en a beaucoup.
[prc‘rzc‘iénotrrsiljdnaboku]
‘Take another, there are a lot of them.’
In the final exercise, learners are provided with 15 sentences in which to transcribe and
label the obligatory, optional, and impossible liaisons. The first two sentences of the
exercise are repeated here (175):
1. Quand est-ce qu ’il aura trente-neuf ans? Le premier avril?

‘When will he be 39? April first?

67

2. [Is se sont acharnés a peindre des arcs-en-ciel.
‘They persisted to paint rainbows.’

APPENDIX C
RATER INSTRUCTIONS
Instructions to the rater".
Take a glance at the following letter and word list. You will be listening to see if the
participant produces liaison between the words in red. Please follow these conventions
for recording your responses to the data presented to you.

1. If the participant produces a liaison consonant, circle the two words and write the
sound they used next to the phrase, as in example (a) and (b).

(a) w for‘beaux hiboux’

or

(b) [bos] hiboux — for ‘beaux hiboux’

If the participant does not produce liaison, do nothing, as in example (c).

(c) [bo] hiboux for ‘beaux hiboux’
You do not need to determine whether or not the liaison is correct or appropriate. You
only need listen to see if the participant produces a liaison consonant between the words
in red.

There are 21 participants and each one has a code. Please write both the code and the
tape title on the rater sheet.

Thank you for your help. Please let me know when you are finished. If you could finish
them in a week, that would be great. I understand that you are busy, so if you need

longer, just tell me.

69

Code Number of Participant:
Tape Title (1a, 1b, 2a, 2b):

 

Chere tante Georgette,

Bonjour! Je suis tres contente d’étre aux Eats-Unis. Le Colorado est si beau. I] y a des
arbres partout. Pres de la maison, il y a an immense jardin. I] ressemble a men ancien
jardin a la Bemerie. La seule difference est qu’il y a des horribles bétes partout. I] y a
beaucoup de fleurs diverses et un lutin ancien ct artisans] au milieu. Mon activité
pre'ferée est de me reposer sur un petit llamas au jardin. J’aime bien faire les magasins
aussi. Ma famille américaine m’a dit que je dépense de l’argent a tort et a travers.

J ’aime bien le travail au pair. Jean at Anne ont quatres enfants. Ils s’appellent Martin,
Manon, Sébastien, et Alice. Je m’occupe d’eux tous les spree-midis sauf le Sarnedi.
Alice a neuf ans, Manon a six ans, Se’bastien a huit ans at Martin a onze ans. Les filles
ont vu des photos de toi et oncle Antoine et elles m’ont dit “Viennent-ils chez nous?”
Elles voudraient bien de faire vos connaissances nez a nez. Sébastien est le comédien de
la famille. I1 aime beaucoup les plaisanteries. Martin, il est coquin. C’est un enfant
insupportable, mais il devient sage petit a petit. Son activité préférée est manger. I]
aime beaucoup les hamburgers et les haricots.

Les enfants aiment beaucoup les animaux. Ils étaient bien excites quand nous sommes
allés au zoo hier Nous avons vu des énormes elephants, des bean: hiboux, des petits
agneaux, et beaucoup d’autres animaux. Manon et Alice ont pris plusieurs photos avec

leurs appareils-photos. Elles avaient plusieurs photos de ces animaux de’ja, mais elles
ont insisté qu’on ne peut jarnais prendre trop de photos.

Enfin, comment vas-tu tante Georgette? Et mes cousins Charles at Monique? Sent-ils a
l’université? Que choisissent—ils comme métier? Et oncle Antoine et oncle Bernard,
travaillent-ils toujours ensemble ou est-cc que oncle Antoine a trouvé d’autre emploi?
Alors, il faut que je parte.

Ecris-moi vite!

A bientot,

Sandrine

7O

11. Please read the following phrases aloud. Please read down each column.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

les apres-midis 1e Colorado des photos au pair

le Colorado pres de les hambupgers Martin a

je parte ct artisanal ecris-moi j’aime bien
choisissent—ils il est autre emploi mes cousins
elles veulent la sernaine at Alice Iutin ancien
me reposer et Anne 51 bientot il aime

nez a nez des plaisanteries plusieurs photos tres contente
nous sommes mes cousins viennent-ils i tort et a travers
je suis Etats-Unis comme métier le travail

mon ancien ils s’appellent chez nous il ressemble
le zoo comme métier des énormes dos horribles
de photos Manon a écris-moi seule difference
beau: hiboux plusieurs photos le travail au milieu
écris-moi la mai son elles avaient ils étaient

il faut ces animaux le Samedi il devient

et elles le travail tante Georgette les filles

1e comédien Alice a enfant insupportable sont-ils

le travail nous avons fleurs diverses la maison
petit apetit il aime je parte si beau

au pair au zoo son activité travaillent-ils
mes cousins les haricots je de'pense le comédien
mon activité famille américaine tante Georgette a dit

les magasins oncle Antoine elles out les magasins
autre emploi petit harnac il faut le zoo

an immense a bientot il devient me reposer

 

7l

 

 

APPENDD( D
TEST FOR PARTICIPANTS

1. Please read the following letter first to yourself. Ask the researcher any questions you
have about the reading selection (vocab, etc.) Now read through the letter twice on tape.

Chére tante Georgette,

Bonjour! Je suis tres contente d’étre aux Etats-Unis. Le Colorado est si beau. 11 y a des
arbres partout. Pres de la maison, il y a un immense jardin. II ressemble 2‘: mon ancien
jardin a la Bemerie. La seule difference est qu’il y a des horribles bétes partout. I] y a
beaucoup de fleurs diverses et un lutin ancien et artisanal au milieu. Mon activité
préferée est de me reposer sur un petit harnac au jardin. J’aime bien faire les magasins
aussi. Ma famille ame'ricaine m’a dit que je dépense de l’argent a tort et a travers.

J’aime bien le travail au pair. Jean et Anne ont quatres enfants. lls s’appellent Martin,
Manon, Sébastien, at Alice. Je m’occupe d’eux tous les apres-midis sauf le Samedi.
Alice a neuf ans, Manon a six ans, Sébastien a huit ans et Martin a onze ans. Les filles
ont vu des photos de toi et oncle Antoine et elles m’ont dit “Viennent-ils chez nous?”
Elles voudraient bien de faire vos connaissances nez a nez. Sébastien est le comédien de
la famille. II aime beaucoup les plaisanteries. Martin, il est coquin. C’est un enfant
insupportable, mais il devient sage petit a petit. Son activité pre'fe'rée est manger. Il aime
beaucoup les hamburgers et les haricots.

Les enfants aiment beaucoup les animaux. lls étaient bien excites quand nous sommes
allés au zoo hier Nous avons vu des énormes elephants, des beaux hiboux, des petits
agneaux, et beaucoup d’autres animaux. Manon et Alice ont pris plusieurs photos avec
leurs appareils-photos. Elles avaient plusieurs photos de ces animaux déja, mais elles ont
insisté qu’on ne peut jamais prendre trop de photos.

Enfin, comment vas-tu tante Georgette? Et mes cousins Charles et Monique? Sont-ils a
l’université? Que choisissent-ils comme métier? Et oncle Antoine et oncle Bernard,
travaillent-ils toujours ensemble ou est-ce que oncle Antoine a trouvé d’autre emploi?
Alors, il faut que je parte.

Ecris-moi vite!

A bientot,

Sandrine

72

H. Please read the following phrases aloud. Please read down each column.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

les apres-midis le Colorado des photos au pair

le Colorado pres de les hamburgers Martin a

je parte et artisanal ecris-moi j’aime bien
choisissent—ils il est autre emploi mes cousins
elles veulent la semaine et Alice lutin ancien
me reposer et Anne a bientot il aime

Nez a nez des plaisanteries plusieurs photos tres contente
nous sommes mes cousins viennent-ils a tort et a travers
je suis Etats-Unis comme métier le travail
Mon ancien ils s’appellent chez nous il ressemble
le zoo comme métier des énormes des horribles
de photos Manon a écris-moi seule difference
beaux hiboux plusieurs photos Ie travail au milieu
écris-moi la maison elles avaient ils étaient

il faut ces animaux le Samedi il devient

et elles le travail tante Georgette les filles

le comédien Alice a enfant insupportable sont-ils

le travail nous avons fleurs diverses la maison
petit a petit il aime je parte si beau

au pair au zoo son activité travaillent—ils
Mes cousins les haricots je dépense le comédien
Mon activité famille américaine tante Georgette a dit

les magasins oncle Antoine elles ont les magasins
autre emploi petit harnac il faut le zoo

uri’ immense a bientot il devient me reposer

 

 

 

“a

 

APPENDIX E

ANONYMOUS QUESTIONNAIRE

The Effects of the Explicit Teaching Method on Oral Production
by Second Language Learners

This is an anonymous questionnaire. For each question, circle the most appropriate
answer given your classroom experiences in the French courses you are taking this
semester.

Questions 1-5 deal with recognizing liaison (i.e. being able to tell when liaison should
occur).

1. In the French 330 Phonetics course, the explanation presented in the textbook
helped me recognize when liaison should occur.

1 ................ 2 ............... 3 ............... 4 ............... 5
Strongly Disagree Neutral Agree Strongly
Disagree Agree

2. In the French 330 Phonetics course, the explanation presented in lecture helped
me recognize when liaison should occur.

1 ............... 2 ............... 3 ............... 4 ............... 5
Strongly Disagree Neutral Agree Strongly
Disagree Agree

3. In the French 330 Phonetics course, the written transcription exercises helped me
recognize when liaison should occur.

1 ............... 2 ............... 3 ............... 4 ............... 5
Strongly Disagree Neutral Agree Strongly
Disagree Agree

74

4. In the French 330 Phonetics course, reading sentences aloud in class helped me
recognize when liaison should occur.

1 ............... 2 ............... 3 ............... 4 ............... 5
Strongly Disagree Neutral Agree Strongly
Disagree Agree

5. Other French courses I am taking this semester helped me recognize when liaison

should occur.
1 ............... 2 ............... 3 ............... 4 ............... 5
Strongly Disagree Neutral Agree Strongly
Disagree Agree

Questions 6—10 deal with producing liaison (i.e. being able to produce liaison correctly in
speech).

6. In the French 330 Phonetics course, the textbook helped me to produce correct

liaison in speaking situations.
1 ............... 2 ............... 3 ............... 4 ............... 5
Strongly Disagree Neutral Agree Strongly
Disagree Agree

7. In the French 330 Phonetics course, the lecture helped me to correctly produce

liaison in speaking situations.
1 ............... 2 ............... 3 ............... 4 ............... 5
Strongly Disagree Neutral Agree Strongly
Disagree Agree

8. In the French 330 Phonetics course, the written transcription exercises helped me
to correctly produce liaison in speaking situations.

1 ............... 2 ............... 3 ............... 4 ............... 5
Strongly Disagree Neutral Agree Strongly
Disagree Agree

75

9. In the French 330 Phonetics course, reading sentences aloud in class helped me
to correctly produce liaison in speaking situations.

1 ............... 2 ............... 3 ............... 4 ............... 5
Strongly Disagree Neutral Agree Strongly
Disagree Agree

10. Other French courses I am taking this semester helped me to correctly produce

liaison in speaking situations.
1 ............... 2 ............... 3 ............... 4 ............... 5
Strongly Disagree Neutral Agree Strongly
Disagree Agree

76

 

ENDNOTES

‘ There are certain glide-initial words that do not permit liaison. For example, mon
whisky, [mfiwiski], ‘my whisky’ and les yaourts, [lejaurrt], ‘the yogurts’ typically do not
occur in liaison while les oiseaux, [lezwazo], ‘the birds’ les yeux, [lezye], ‘the eyes’ do.

H-aspiré words form an additional exception to this rule of liaison occurrence. For

example, the phonetic representation for les hamburgers is [ledburrgrerr] and not
[lezdburrgcers].

2 Valdman represents the underlying form of liaison consonants by using a capital letter
(Valdman, 1976: 102).

3 Note that in this case, the vowel denasalizes. This phenomenon occurs in numerous
prenominal adjectives in standard French (Tranel 1987:84).

4 Notation conventions have been altered in order to be consistent with the conventions
used in (4).

5 The nasal vowel [ti] seems to be in the process of merging with [E] in most varieties of
French (Ayres-Bennett et al: 2001). It is presently common to see [(2] and [E] used
interchangeably in phonetic transcriptions that have traditionally used [(2].

6 Personal pronouns form a subset of general pronouns and so accounts that use the
general term ‘pronoun’ are assumed to also include ‘personal pronouns’. This is
consistent with the examples these researchers provided in their analyses.

7 These are adverbs that are not listed in Table 2.6.

 

8 These are prepositions that are not listed in Table 2.6.

9 Tranel and Dumenil make an exception for Comment allez-vous? ‘How are you?’,
which they treat as a case of obligatory liaison.

1" An exception to this rule occurs with _ et un ans/heures. See table 2.6.

" An exception to this rule occurs when a determiner combines with an h-aspiré word as
in des hiboux [deibu] ‘sorne owls’.

‘2 See Appendix A for a breakdown of background information for participants.

‘3 See Appendix B for input for explicit instruction.

'4 The liaisons mentioned in passing fell under the category of determiner-noun and were
mentioned while the class was working on transcriptions focusing on another topic.

‘5 See Appendix C for rater instructions.

‘6 See Appendix D for the test medium.

‘7 Set phrases were chosen directly from the textbook used for instruction.

‘8 H-aspiré words were chosen directly from the textbook input used for instruction.

‘9 Although l’Académie Francaise has recently recognized this liaison as appropriate, it is
still treated as an impossible context by many speakers as well as textbooks and curricula.
2" Set phrases were chosen directly from the textbook used for instruction.

2‘ See Appendix E for the questionnaire.

2‘ See Appendix D for task instructions.

2’ An exception to this rule occurs when a determiner combines with an h-aspiré word.

2“ See Appendix E for questionnaire.

2’ This question only applied to the IMPINS group.

2" This question only applied to the IMPINS group.

78

 

‘7 Note that this is really a case of encha’r‘nement and assimilation and not liaison. See

page five for further explanation.

79

 

REFERENCES

Ayres-Bennett, Wendy, Janice Carruthers, and Rosalind Temple. 2001. Problems and
perspectives: studies in the modern French language. Essex: Pearson Education Limited.

Battye, Adrian, Marie-Anne Hintze, and Paul Rowlett. 2000. The French language today:
a linguistic introduction. New York, NY: Routledge.

Booij, Geert, and Daan De Jong. 1987. The domain of liaison: theories and data".
Linguistics. 25(5). 1005-1025.

Clements, George N. and Samuel J. Keyser. 1983. CV Phonology: A generative theory of
the syllable, Cambridge, MA: MIT Press.

Culicover, Peter W. 1982. Syntax. New York, NY: Academic Press, Inc.

Delattre, Pierre. 1966. Studies in French and comparative phonetics. Netherlands:
Mouton & Co.

Dell, Francois 1973. Les Regles et les sons: introduction a la phonologie generative.
Paris: Hermann.

Dumenil, Annie. 2003. Facile a dire! Les sons du francais. Upper Saddle River, NJ:
Prentice-Hall.

Ellis, Nick. 1994. Implicit and explicit language learning-An overview. Implicit and
eXplicit learning of languages, ed. by N. Ellis, 1-31. London: Academic Press.

Gass, Susan M. and Larry Selinker. 2001. Second language acquisition: An introductory
course. Mahwah, NJ: Lawrence Frlbaum Associates, Inc.

Gaudin, Lois S. 1953. Common mistakes in pronunciation. The French Review. 26(6).
451-460.

Green, Peter and Kariheinz Hecht. 1992. Implicit and explicit grammar. Applied
Linguistics 13(2).168-184

Haegeman, Liliane and Jacqueline Guéron. 1999. English grammar: A generative
perspective. Oxford: Blackwell Publishers Ltd.

Hawkins, Roger. 2001. Second language syntax: a generative introduction. Oxford:
Blackwell Publishers Ltd.

 

 

Hulstijn, Jan. 2002. Towards a unified account of the representation, processing, and
acquisition of second language knowledge. Second Language Research. 18(3). 193-223.

Klausenburger, Jurgen. 1974. Rule inversion, opacity, conspiracies: French liaison and
elision. Lingua. 34.167-179.

Klausenburger, Jurgen. 1984. French liaison and linguistic theory. Stuttgart: Steiner.

Labov, William. 1972. Sociolinguistic Patterns. Philadelphia, PA: University of
Pennsylvania Press, Inc.

MacWhinney, Brian. 1997. Implicit and explicit processes. Studies in second language
acquisition. 19.277-281.

Olsen, Carroll L. 1973. A course in corrective phonetics. The French Review.
XLVI(3).573-583.

Prunet, Jean-Francois. 1986. Studies in Romance Languages, ed. by Carol Neidle and
Rafael A. Nur'iez Cedefio, 25.225-235. Dordecht: Foris Publications Holland.

Schane, Sanford A. 1972. French phonology and morphology. Cambridge, MA: MIT
Press.

Selkirk, Elisabeth. 1974. Phonology and syntax : the relation between sound and
structure. Linguistic Inquiry. 5(4).573-590.

Thomas, Alain. 1998. Liaison and liaison instructibn: from pronunciation models to
linguistic reality. The Canadian Modern Language Review/La Revue canadienne des
langues vivantes. 54(4).543-552.

Tranel, Bernard. 1981. Concreteness in Generative Phonology: evidence from French.
Berkeley, CA-London: University of California Press.

Tranel, Bernard. 1987. The sounds of French. Cambridge: Cambridge University Press.

Tranel, Bernard. 1995. French final consonants and non-linear phonology. Lingua.
95.131-167.

Trudgill, Peter. 2000. Sociolinguistic: An introduction to language and society. London,
England: Penguin Books.

Valdman, Albert. 1976. Introduction to French phonology and morphology. Rowley,
MA: Newbury Hous Publishers, Inc.

81

 

 

Walz, Joel. 1980. An empirical study of pronunciation errors in French. The French
Review. 53(3).424-432.

82