ON THE STRUCTURAL UNITY OF THE AKAN

DIRGE

by E.  O. Apronti

Introduction

An important  aspect of the  total study of  Oral  Literature  in Africa  relates to  the
inner  dynamics  of  the  texts  as  texts.  Such  studies  complement  what  can  be  stated
about  the  religious,  social or  ritual  contexts in which  the texts  are performed.  They
do  this  principally  in  increasing  our  awareness  of  the  factors  that  have  fostered  the
survival of these texts as texts, while at the same time making explicit those  structural
dynamics that  reinforce  the literary  appeal of the texts.

Preservation  and  survival  are  crucial  factors  in  the  "lives"  of  unwritten  texts.
These texts vary in the extent to which it is felt that they should retain their  integrity.
The "prose" narrative of a folk tale for  instance allows for more textual  improvisation
than  does  the  Yoruba  babalawo's  IFA  "verse"  chant.  Generally,  indeed,  the  per-
formance  of verse texts tends to require a greater fidelity  to the original than  does the
performance  of prose texts.

How, then, does the integrity of an unwritten  text get preserved?  Content-wise,
most texts evince a unity in their structure.  Not only are they concerned with a parti-
cular  topic, they  also generally  draw on vocabulary  from  a set field  or scatter.  They
thus evince a coherence of theme or of  subject.

Structurally,  also,  they  evince  various  devices  which  unify  the  texts  and  thus
enhance  their  memorability.  The  extant  literature  on  oral literature  high-lights  this
structural  feature;  a  few  examples  are  Babalola's  Content and  Form of  the  Yoruba
Ijala,  OUP,  1966; Abimbola's  "Stylistic  Repetition  in  Ifa  Divination  Poetry," Lagos
Notes and Records III,  I,  1971, 38-53,  and  Nketia's  Funeral  Dirges  of  the  Akan,
Achimota,  1955, pp. 75-112.

Structural unity in both its syntactic and lexical aspects can, therefore,  be shown
to be a characteristic of many oral literature texts.  As has been suggested  above, such
structural  unity  is a function  of the  memorability  of  texts—the  tighter  the  structural
dynamics, the greater the chances of a text remaining unchanged; conversely the looser
its structural unity, the more liable the text is to emendation.

The Akan Dirge

The  most  outstanding  study  of  this  literary  form  is  Professor  Nketia's  book  to
which reference  has been made above.  In it is to be found  a succint discussion of the
religious  and  social  contexts  in  which  dirges  are  performed,  an  analysis  of  their  lin-
guistic and literary properties, and examples of a variety of dirge texts.

*  Dr.  E.  O. Apronti  is a  Research Fellow in  Linguistics.

32

What is intended here is to supplement, without undue repetition, the statements
that  Nketia  makes  on  the  structural  unity  of  the  dirge.  The  theory  which  we  shall
apply to this material is a version of the concept of COUPLING  proposed  by Samuel
R. Levin in his Linguistic Structures in Poetry, Mouton, The Hague, 1962.  The theory
is applicable  in  all its rigour  to the  dirge texts to  be discussed.  We desist  from  this,
however,  because such an exercise will interest only specialists; we are here concerned
with  the  type  of  statements  that  readers  who  are  not  specialist  linguists  would  find
congenial. 

. . ..

Poetry  versus Prose

Before  expounding the outline, of Levin's theory  and illustrating its applicability
to the Akan dirge, it might be useful  to dispose of the problem of distinguishing verse
or poetry  from  prose.  This, of course, has been a subject  of age-long debate.

The  distinction is usually, but  by no means always, clear-cut.  Indeed  the dirge,
being a song text, is poetry almost by definition.  But the question raises wider stylistic
issues of which we must dispose, if only to justify  the  relevance  of  Levin's  theory to
our  discussion.

This distinction is made difficult  by the fact that some "poems" are clearly prosaic
in  flavour  (which  generally  makes  them  bad  poetry.  Examples  abound  in  the  daily
and  weekly  press).  Even  more  disconcerting  is the  tendency  of  some  prose  texts  to
evoke in all but  their form  the responses that readers generally associate with  poetry.
A recent example is afforded  by the small-print chapters of Kofi Awoonor's first novel,
This Earth, My  Brother

Attempts  that  scholars  have made to  distinguish  poetry  from  prose  have  varied
from  appeals  to  memorability,  to obviously  regular  rhythms, to  structural  symmetry
and  to  the  congruence  of  'form'  and  'content.'  These  criteria  sound  formal  and
rigorous enough, until it is remembered that many prose statements or passages exhibit
them as well.

Several recurrent quotations undoubtedly have a prose form. For instance, an ultra-
colonialist  statement  by Britain's war-time leader  Sir Winston  Churchill  to the  effect
that he would not preside over the liquidation of the British Empire has found  echoes
in many parts of the English-speaking world.  Besides, proverb texts which are quoted
to such good effect  in ordinary life can also be said to be prose.  But they often  evince
structural  symmetry.

Such  symmetry  is  also  evident  in  the  prose  style  called  "Eupheuistic  writing"
once popular among English prose writers and employed also by J. E.  Casely-Hayford
in his pseudo-novel, Ethiopia Unbound (1911).

It  would  seem  that  the  most  satisfactory  clues in  these  days  of  "prose  poems,"
"free  verse"  and  "concrete  poetry"  consist  of  a  combination  of  these  various  traits.
We  can  thus  make  explicit  our  sensation  of  a  structural  unity  in  poetry  principally
on the basis of the syntax of the poems and secondly by reference to its meaning and
the peculiar  organisation.of  its (poetic) expression.

It  will be  found  that  the longer the piece, the more fruitful  it is to analyse it on
these structural  lines.  Indeed when prose—even fine poetic  prose—parades  as verse

33

the  structural-syntactic  test  readily  shows  it up  as  not  being  as  closely-knit  as  we
would  expect  a  piece  verse  to  be.  The  kind  of  structural  labels  we shall  provide
for  the  dirges below are clearly inapplicable to prose texts.

Levin  and  "Coupling"

The  attraction  of  Levin's theory  is that  it  helps to  elaborate the  structural  basis
of  Nketia's  stylistic  analysis.  It  is  hoped  that  this  will  enable  us  to  show  that  the
form  and  the  content  of  a traditional  oral poem  can  evince as much symmetry  as an
individual composition (compare the discussion offered  below with Jean  Ure's  stylistic
analysis  of  the  present  writer's  poem  "Funeral,"  in  Proceedings  of  the Eighth  West
African  Languages Congress, University  of  Abidjan,  1971,  Volume  2,  p.  589-597).
To  summarize  Levin's theory, he claims that the unity or symmetry that  a poem
evokes  is a function  of  a particular  type  of  composition  or  structure.  He  recognizes
various units of structure, including the line.  These units have places in them, which
are  occupied  by  smaller  (linguistic)  units.  These  latter  units,  naturally,  have  inter-
relations both within the line and in successive lines.

The first  line of a poem may consist of a structure of subject plus Verb (transitive)
plus  Object.  The  next  line, let  us say, may  exhibit  a similar  structure.  This  is one
kind  of  symmetry.  The  two subject  positions are said  to  be equivalent  (i.e. similar).
But then,  factors  on other  linguistic planes  (apart  from  grammar)  may reinforce  this
symmetry.  For  instance  the  respective  lexical  items  (or  items  of  vocabulary)  that
occupy  Subject  place  in  the  two  hypothetical  lines  may  be  semantically  congruent
(where for  instance  they  are synonyms),  or  otherwise.  Similarly,  the  lexical  content
of the two verbs may correspond  or contrast and so on.

The  possibilities  are  therefore  quite  considerable.  One  can  pursue  this  type  of
analysis, on the one hand, by reference to such linguistic levels as phonology, morpho-
logy, syntax etc., and, on the other hand, in relation to combinations of varying numbers
of lines.

Levin  draws  a  distinction  between  positionally  (i.e.  syntactically)  equivalent
classes  which  he  calls  Type  One,  and  semantically  equivalent  classes,  Type  Two.
The ideal of structural unity and symmetry, then, occurs when Type Two equivalences
coincide with (or are embedded in) Type One equivalences1  Then you have 'coupling.'
That  is,  we  have  coupling  when  linguistic  items  which  are  either  semantically  (or
phonically) equivalent are found occurring in comparable structural positions.

The  full  rigour  of  this  theory  invokes  the  phonology  or  sound  system  of  the
language  in  which  the  poem  occurs.  As we  shall  be  dealing  here  with  the  English
translations of the  Akan  dirge, it would be fruitless  to pursue the full  implications of
this aspect of the theory.  Such  an exercise would  reveal, in the phonological  sphere,
only fortuitous phonic  equivalences  in  the  English  text.  This  is not  to  say that  the
latter  are entirely  devoid of interest.

But  the first  stage of the theory which  deals with  syntactic and  semantic  factors,
we contend, sheds a useful  light on the structure of the Akan dirge, to which we now
turn.

34

The Structure  of the Dirge

One aspect  of linguistic analysis which will feature  in the discussion that  follows
is the  revelation  of  the  skeletal  structure  of  linguistic  matter.  To  do  this,  we  need
symbols  that  can  extrapolate  that  structure  and  abstract  it,  thus  high-lighting  struc-
tural  features  (namely  correspondences  and  contrasts)  to  the  exclusion  of  all  other
features.

We propose to use the following symbols:

Z 

•—  addressive,  similar  to  the  Latin  vocative,  though  not  formally
marked  in  all  the  texts.  (One  of  the  few  marked  is  "Asumegya
Gyebirioo"p.211, line 6).

S  —  Subject,  defined  by its  regular  syntactic  relationship  to  a  Verb
or  Predicator.  It  may  in  addition  precede  or  follow  forms  of  the
copular, the verb "to be."

P  —  Predicator or verb, evincing tense and mode forms,  and in syntactic

relationship to a nominal group in subject position.

C  —  Complement  or  direct  Object,  a  nominal  group  dominated  by  a

transitive verbal group.

Q 

R 

•—  Indirect  objects.

•—  locationals,  separated  from  other  adverbial  phrases  (A,  below)  on
account  of the  dominance  of the theme of the  place of domicile in
the  dirge  (see  Nketia,  op 

cit  ,  pp.  38-43).

A 

•—  adverbials other than locationals (R, above).

To turn now to some of the longer dirge texts which, as we have already indicated,
yield the most fruitful  results through this type of analysis.  The first Aduana dirge on
page  139 begins:

1.  Atwea Yaa, the flint  arrived too late.

Here "Atwea  Yaa"  is  an  addressive  or  appellation,  and  is  therefore  "Z".  The
next  two words constitute a nominal group  " S" in subject  position  to the  Predicator
"P"  "arrived." The last two words being adverbial (but not locational) may be labelled
"A."

Line One therefore  yields the structure:

1.  Z,  S  P  A

Line Two reads:

2.  Sakrabutu, he is neither an upturner, nor one who moves about.

35

We open  again  with  an  addressive  "Z," followed  by  two  clauses  which,  for  our  pur-
poses,  ought  to  be  rewritten  (because  the  following  are  their  underlying  strings  or
structures)  as 

V 

V

2a  he  is  not  an  up  turner 

' 

' 

' 

2b  he does not  move about. 

A •'•:'

'

It is interesting  here  that  the  Akan  version  (page  209,  line  2)  evinces  a  structural
symmetry which is obscured by the English  translation:

2c  Sakrabutu,  onye butufoo,  anyE nantefoo.

which is clearly Z,  SPS,  ,$PS,, where P is the copular  "to  be" to  which the  category
of transitivity  does not  apply, hehce,i'S'after'P.' 

. 

,

To  turn  back to the  English version,  then,  the structures  are:

2a  S  P  S

2b  S  P  R

We may  note  here  in  passing  the  uniform  negative  mode  of  the  two verbs, as well as
their uniform  tense.  Line 2 as whole thus yields the structure:

2.  Z,  S  P  S, 

S  P  R

"\  The  next line reads: 

;

3.  You will find him crouching by the wayside

which we may expound as

3a.  You will find him

3b.  He is crouching by the wayside. 

These two yield the following  structures

• ' ' , • :•

3 a. 

S  P  C 

• 

• i 

-• 

.  .,, 

.- 

;

Where  "you"  is  " S ," "will  find"  is "P," and  "him"  is  "C."

3

b. 

S 

P 

R 

.- 

• 

• . ' . ' . . .' . ' . ' . . . , : - ' , ', 

."  ' • • '-  ••••••'•• 

- 

.

'!- " '

•

' ? -•

With  "he" as " S ," "is  crouching"  as "P," and  "by the  wayside" as "R"  a locational.
Thus the whole line is: 

; :  *  •

'' 

3. 

S  P  C, 

S  P  R 

•

The  next two lines embody a conditional  clause 

•
If you  ask him for  anything,  he gives it to you.

4. 

36

5.  And  says:  Vanquish the thousand and the mighty with it,

which  is balanced by lines 6 and 7.  Line 4 yields the structure 

..

4.  S  P  C  Q, 

S  P  C  Q

an  example  of  perfect  symmetry  within  the  line, a fact  that  is not  readily  evident  in
the verbalized  version.  Line 5 may be said to be prefaced  by  "if  you  asked him  for
anything"  (SPCQ) of line 4; 

\-\  ... ;

"He  says" can therefore  precede the colon as SP, thus:

5. 

( S P C Q,  S)  P:  ';''  P C* V  C? A

We have here  a  concatenation  of  two  direct  objects  dominated  by  the  verb
"vanquish,"  namely "the thousand" and "the mighty."  Since we cannot assume that
those  two  have  identical  referents,  we'use superscript  numbers to distinguish  them.

Now  come the two lines that  counterpoise the previous two;

6.  Then  you reply:  I will not use it for  that purpose.

7.  But will use it for something greater.

Line 7 in its i;urn is expounded by lines 8 atid 9'; as we shaft1 see below later.- •
,

Line 6 yields the structure 

. 

6.  S  P: 

S  P  C  A

Line 7  begins  with  a  suppressed  "Then  you  reply,  (But)  I  " . ..  .which we supply,

y

i

e

l

d

i

ng 

' 

:" 

;" 

7. 

•  "  "•- 

(S  P:  S) 

'' 

' 

'•• 

'

•< 

< 

""•''- 

•••'••• 

- 

•• 

-" 

• 

•  • 

•'• 

• • •' 

••

P  C  A

The symmetry  evinced between lines 6 and 7 is thus made explicit; We note  in  addi-
tion the  contrasting modes or poles of the two verbs, a negative "will not use" in line
6, but  an  affirmative  "will use" in line 7.

The  next two lines read:

8.  Deception and the worst  stratagem
9. The  portion  of the exterior of the pot  that  Skin does not  cover.
Line  8  is  structurally  dominated  by  "I'wilt  use it  f6f"  in  line 7  and  is thus  an

elaboration  of the "A" element of that line.  It therefore  has the  structure

8.  A1  +  A1 
Line 9 might  be rewritten:

£A  -1  r.\

9a.  The portion of the exterior of the^ pot
9b.  Skin does not cover it

where 9 is structurally similar to 8, hence

and 9b yields

9a.  As

9b.  S  P  C

The complete line is thus

9a.  A»,  S  P  C

The final line is

10.  Grandchild  of  Atwea  Yaa  that  hails  from  the  town  of  Atwea

Gyampon.

which we may re-write as 

yielding 

10a.  Grandchild  of Atwea Yaa 
10b.  You hail from the town of Atwea Gyampon,
10a.  Z
10b.  S  P  C

,;f'i 

-  - *•'

' , ,r  -

The transitivity  of the Predicator  "hail from"  (equivalent to  the Akan  "firi")  is veri-
fiable by  replacing  "the  town  of  Atwea Oyampon"  with  the  pronoun  object "him."
"From.  . . ..  .Gyampan" is, thus not a locational unit,  despite all appearances to  the
contrary.

Line 10 as a whole thus yields:

10.  Z, 

S  P  C

The skeletal framework  of the first Aduana dirge on Nketia's  page 139 may  thus

be displayed  as:

S  P  R

S  P  R

S P CQ

PC1  +  C?A

1.  Z,  S  P  A
2.  Z,  S  P  S, 
3.  S  P  C, 
4.  S P C Q, 
5. 
6.  S  P: 
7. 
(S  P:  S) 
8.  A1  +  A*
9.  A3,  S  P  C
10.  Z, 
S  P  C

(S  P  C  Q,  S)  P: 
S  P  C  A

P  C  A

38

Recurrent  syntactic  structures  are  thus  seen  to  be  a  significant  feature, of  the
dirge.  Add to this the consistency of the vocabulary, drawn here from  the activity of
hunting and combat:  flint,  moves about, crouching, vanquish, deception, stratagem,  skin.
We  thus  get  an  inkling  of  the  basis  of  the  reader's  sensation  of  unified  or  coherent
experience in the dirge as a form of poetry.

We  conclude  the  paper  by  citing  another  dirge  and  extrapolating  its  structure
to  illustrate  the  method  employed  here.  It  is of course possible to specify  in  greater
detail  than  we  have  done  in  this  paper  the  structural  elements  that  underline  the
unity  of the  dirge, but  we have indicated  above reasons why that  amount  of  delicacy
(i.e. detail in analysis) would be inappropriate  here. In  the following dirge, therefore,
many locational references will be adumbrated by a "Z" addressive specification.  This
is  done  without  prejudice  to  further  statements  that  are  possible  and  are, indeed, of
interest,  but  which we shall not accommodate  here.  (Note  again that  "hail  from"  is
treated  as a transitive verb).

The  dirge in question is number  90 on Nketia's page  182.  It  is an Agona  dirge
1.  Kotoku person and grandchild of the Vanguard of Kotoku.

2.  Grandchild  of  Ampoma:  our lineage hails from  Kotoku

Z 

, 

S PC

3.  Grandchild  of  Baabu:  our lineage hails from  Kade

Z 

, 

S PC

4.  Grandchild  of Ofori  Amanfo:  you are a Tia person

Z 

, 

S PS

5.  Grandchild  of  Ampona:  our  lineage  hails  from  Dankyira  Mmorebemu.

Z 

, 

S PC

6.  Grandchild of Sasu Dam hails from Sankubanase (when rewritten as "Grand-

child  of  Sasu  Dam, you hail from  Sankubanase).

Z 

, 

S PC

7.  It  is grandsire  that  hails from  Kontonkyi, where  the stone  is wearing  down

the axe

S P S, 

S P C, 

R S PC

8.  Tia  person,  grandchild  of Akwasi Afire

Z 

Z

, 

39

9.  Okyeregya's grandchild that drinks at Tafo Aboaso 

: >,.; 

;

• 

- 

' 

• 

Z 

, 

S P R* 

. 

• .Wr 

u ' : . . - . . ' - ) ~ . - ' \ 'l

10.  Abeam Dansoaa offspring  of the Biretuo clan of Sekyere. 

'•••  T

• 

•' 

' 

• 

' ' .1 

'••" 

• 

) 

'

.» 

;" 

'

Z 

, 

Z

It.  Afcwasi  Afre  is the offspring  of Agona  and Biretuo  clans

S PS

12.  A Dompre  person  and grandchild  of Aboagye  Awua of  Akyem

Z 

Z

13.  Grandchild  of Abora  Kani,  Kani pure  and true

Z 

, 

Z

•  or, following  Akan  syntax,  Z,  S P C.

Apart,  again,  from  the closely-knit  structure  of  syntactical  elements,  we  notice
even  from  a casual  glance the elaborate  repetitions  of kinship  vocabulary'—grandchild,
offspring,  hail,  lineage  This  dirge is therefore  a unified  poetical  experience.

The  case for regarding  the dirge as a coherent  whole is thus  seen to be based  on
structural,  lexical  and semantic  considerations.  We  have  been  concerned  here  with
drawing  the  outline  of  the application  of  a  method  of  stylistic  analysis  which  lends
validity  to  our  intuitive  awareness  of the  appeal  or  "punch"  of  a  dirge  as a  unified
poetic expression, and we find here at least a partial explanation as to why, even without
the  benefit  of writing, this piece of oral literature has hung together  for so long.

40