TH3 LANGUAGES OP  TEH! AKM PEOPLES*

By

Florence  Abena  Dolphyne

There  a re various  groups  of  people  who call  themselves  "Akan"»

They  inhabit  most  of  the area  t h at  covers  t he forest  and coastal  b e l ts

bounded  in  the west  by t he Bandama  River  in t he Ivory  Coast  and in t he

eagt  by t he Volta  River  in  Ghana,  and include  t he Anyi,  t he Baule? t he

Nzema?  t he Ahanta,  t he Fante?  the Asante,  t he Akuapan,  t he Akyem, t he

Brong,  e tc  All of  than  seem  to  share  certain  common  cultural 

t r a i ts

that  distinguish  them  from  other  ethnic  groups  in T/est  Africa*

Linguists  (Greenberg,  1963?  and Westermann  .and  Bryan,  1952)  have

used  the name  "Akan"  to  refer  to  t he  group  of  languages  spoken  in t he

geographical  area  r e f e r r ed  to  above,  so  that  t h e ir  u se of t he name

"Akan"  more  or  l e ss  coincides  with  t he ethnographic  use of  t he name*

Greenberg' s  <*Akan"  language  group  is  made  up of  s ix different 

languages

which  a re Anyi,  Baule,  Twi, Guangs  Metyibo  and Abure,  while  Westermann

and  Bryan1 s is  made  up of  t he Twi-Fante  dialect  clusters  the Anyi-Baule

d i a l e ct  clusters  and the Guang  d i a l e ct  cluster..

In  t he 1950's  t he name  KAkan»  was o f f i c i a l ly  adopted  in  Ghana as

the  name  for the language  whose  written  d i a l e c ts  a re Asante,  Akuapem

*This  paper  was written  in  1975  for t he  "Kwame  Da-aku  Memorial  Volume*1*  -
a  volume  dedicated  to t he memory  of  t he l a te  h i s t o r i a n,  Dr. K°Y»  Daaku,
which;  for  various  reasons,  ••• could  not be published^  The  paper  is of
general  interest,  although  the  latter  part  of  it  has particular  interest
for  the  professional  linguist.

and  Fante»  The  well-known  Twi  scholar,  J.G*  Ohristaller,  had  called

t h is  language  Twi,  so  that  both  his  Dictionary  and  Grammar  are  of  fThe

Asante  and  Fante  language  called  Tshi1 *  This  name  was  however  not

acceptable  to  Fante  speakers?  since  Twi  was  generally  understood  to

include  such  dialects  as  Asante9  Akyem,  Akuapem  and Kwahu,  but  to

exclude  Fante«  The  name  wAkanrt  was*  on  the  other  hand,  neutral  enough'

to  be  acceptable  to  both  Fante  and Twi  speakers,  but  i ts  use  in  t h is

limited  sense  was  at  variance  with  i ts  use  in  linguistic  literature*

In  recent  years  Stewart  has  Introduced  the  name Volta-Qomoe  for  the

group  of  languages  called  Akan  by  Greenberg  and  by  Westermann  and  Bryan,

and  he  reserves  the  name  wAkan!t  for  the  Twi-Fante  language,  in  conformity

with  current  usage  in  Ghana*  The present  writer  has  suggested  Volta-

Bandama  as  a  more  appropriate  label  (Dolphyne  1974)  since  the  western

border  of  the  language  group  is  the  Bandama River  and  not  Comoe River.

For  a ll  these  linguists?  the  Akan,  Volta-Comoe''or  Vblta-Bandama

language  family  includes  languages  such  as  Guang,  whose  speakers  are

not  generally  considered  as  belonging  to  the  Akan  ethnic  groups*  It  is

therefore  necessary  that,  at  this  point,  a  clear  statement  is  made  as

to  which  group  or  groups  of  people  are  being  referred  to  in  this  dis-

cussion  as  'Akan  Peoples'*

In  a  paper  'The  Origins  of  the  Akan' ,  A*A» Boahen  said  among

other  things;

The  Akan  is  those  groups  of  peoples  in  Ghana  among
whom  a number  of  identical  or  virtually  identical
cultural  t r a i ts  and  institutions  are  found.  The
first  of  these  unique  t r a i ts  is  linguistic.  All  the
Akan  groups  speak  the  T'?/i  language,  or  dialects  of
the  saine  whi ch  .ar e mutual ly  int el 1 i gib 1 e»°° ° •  The
second  common  t r a it  is  institutional.  All  the  Akan
groups  have  a  common  calendar,  common  religious
beliefs?  naming  ceremonies;  marriage  institutions
and  matrilineal  systems  of  inheritance.  They  also
have  a  monarchical  system  of  government"

In  many  ways?  the  most  interesting  of  these  common
institutions  are  their  identical  exogamous  matrilineal
and patrilineal  clans*  "Each  genuine  Akan  group  is  :
divided  into  eight  principal  matrilineal  and  eight
patrilineal  clans*

On  the  basis  of  these  criteria  he  concluded  that  'the  Akan  are

today  the  Asante,  the  Akyem,  the  Kwahu,  the  Akwapim,  the  Wassa,  the

Twifu?  the  Assin*  the  Panto  and  the  Gomua,' -  All  these  people  are  found

in  Ghana..

In  a  more  recent  paper,  (A«A»  Boahen,  -|974)  ']3ho  are  the  Akan?' 9

the  same  writer  has  a  much  longer  l i st  of  peoples  who  qualify  to  be

called  Akan.  In  addition  to  the  above  l i s t,  he-has ?  among the  Gh-3i a

Akanss  the  Bono,  the  Akwamu,  the  Bueitt,  the  .v.a:onas  the  Sehwi,  the  Aowin,

the  Nzema. and. the  Ahanta» 

In  the  Ivory  Coasts  the  Akan  peoples  are  the

Abron,  the  Anyis  the  Sanv/i,  and  the  Baule*  He  s Ts  that  although  the

cluster  of  peoples  known  n,g  the  Lagoon  peoples  or  Les  Lagunaires*  namely,

the  Atie,  the  Abe,  the  Abides  the  Ad jukuru ,  the  "Sfbrie?  the  Ega  (Die)

the  Eotiles  the  Abure,  the  Avikam  and  the  Alladjan  also  call  themselves

Akan s  he  does not  consider  them  as  such  since  they  do not  share  the

cultural  t r a i ts  which  are  so  characteristic  of  the  Akan«  Of  these

peoples  he  says: 

'Their  political  organisation  is  certainly  a  non-

centralised  one; 

their  week  consists  of  six  days  and  not  seven,  and

their  system  of  naming  is  different,  and  finally  it  does  not  appear  that

the  matrilineal  and  patrilineal  clan  groups  are  found  among  them1 .

It  therefore  seems  that  ?/hen  the  linguist! c5  cultural  and  socio-

logical  evidence  is  taken  into  con si deration  the  conclusion  that  one

may  come to  is  that  the  Akan  peoples  are  those  who  speak' that  group  of

languages  referred  to  by  Stewart  as  Tano5  a  subgroup  of  his  Volta-Gomoe

(yolta-Bandama)  language  family.  These  are  people  who  speak  Baul.e?  Anyis

Sanwi  (Afema),  Nzema,  Atlanta,  Aowin,  Sehwi,  Wassa,  Brong/Abron,  Pante9

Gomua?  Agonas  Asante,  /skyem,  Kwahu  and  Akuapem.  It  is  these  people  who

also  have  in  common  those  cultural  t r a i ts  which  Bo ah en  uses  as  part  of

his  criteria  for  determining  which  groups  of  people  are  Akan»

*The  basis  on  which  Bo ah en  includes  the  Buem  in  this  l i st  is  not  clear.
Linguist ically s however,  the  language  of  the  Buemj  Lelemi?  does  not
belong  to  the  Yolta-Bandama  language  family«  Some  Lelemi  speakers
speak  Akan  (Twi-Pante)  as  a  second  language,  but  this  not  true  of  all
of  them*

It  may  be  of  interest  to  point  out  here  that  although  all  Akan

peoples;  as  defined  above?  have  a  common  system  of  naming  children  in

which  the  f i r st  name  is  derived  from  the  name  of  the  day  of  the  week

on  which  the  child  is  born,  among  speakers  of  Anyi,  Baule  and  Fzema

the  names  for  the  days  of  the  week  and  the  corresponding  names  given

to  persons  born  on  these  days  are  one  day  out  of  step  with  those  used

"by-speakers  of  Akan  (Twi-Fante)-  This  means  that  the  name  for  Monday

(Kinzile/Kissie)  in  Anyi ,  Baule  and  Nzema  corresponds  to  the  name  for

Sunday  (Kwasfe)  in  Akan  and  so  on  as  i l l u s t r a t ed  below. 

(The  names

given  to  persons  born  on  the  days  of  the  week  are  marked  M for 

'male'

and  F  for 

'female')-

ANYI-BAULE

AKAN  (TWI-FANTE)

Monday 

K i n z i l e / K i s s i e:  Kwasi/Kasi 

•; Akisi/Akasi- 

(M) 
( F) 

DZW/DZJUO: 

~ 

Kwadzo/Kodzo
Adzua

Tuesday 

Dzoke/Dzols

Kwadsjo/Kad^o/Kodzo (M)
(F)
kdjuba/Adgua 

Bina 

:

Kwabina
Abina

Kwaku
Akua

:

:

:

:

Yaw/Ekv;aw
Yaa/Aba

Kofi
Afua

Kwand
Ama

Kwasi
Ako sua/Esi

(M)

(M)
(F)

(M)
(F)

(M)
(F)

(M)
(F)

(M)
(F)

(M)
00

Wednesday  Maali/Mla

Thursday  Kule/Uhue

Uwue

Friday 

Yaale/Ya

Saturday 

Fule/Fue

Sunday 

Mule

Kwamla/Kamla
Abina/Abinima
Abla

Kwaku/Kaku
Akuba/Ahu

Yaw
Yaba/Aya

Kofi  :
Afiba/Afua

Kwami
Ama/Amue

Wukuw/Wukuo:

(M)

(F)

(M)

Yaw 

(M)

Fi 

(M)
(F)

(*O
(F)

Mimin 

Kwasie 

Although  speakers  of  Nzema  and  Aowin  in  Ghana.use  the  same  names

for  t h e. days  of  the  week  as  thos-:-  used'by  speakers  of  .Anyi  and  Baule?

the  names  they  give  to  children  born  on  these  days  correspond  to  those

tised  by  speakers  of  Akan  (Twi-Fante)  • so  that  for,  them  there  is  no

correspondence  between  the  names  of  the  days  of  the  week  and  the  names

given  to  children  born  on  those  days»  For  example?  the  Nzema  names

given  to. children  born  on  Thursday  are  Yaw  and  Yaba  which  correspond

to  the  Nzema  day  name  for  Friday,  Yaale,  as  i l l u s t r a t ed  below*

GHANA  NZEMA

Monday

Kinzile

Tuesday

Daoke

Wednesday

Maali

Thursday

Rule

Friday

Yaals

Saturday

Fuls

Sunday

Mule

Kodzo
Adzuba

Kamla/Kabina
Abinliina

R.aku
Akuba

Yaw Awaw
Yaba

Kofi
Afiba

Kwami
Ama

Kwasi
Akasi

(M)

(M)
(F)

(M)

(M)
(F)

00

(M)
(F)

(M)

It  seems  to  me that  this  is  due  to  the  proximity  of  the  Ghana

Nzema  and  Aov/in  (Anyi)  to  speakers  of  Aknn  (Twi-Praite),  who.form. a.

large  linguistic  group  in  Ghana?  and  whose  language  has  tended  to  have

prestige  status  in  part s  of  the  ^zema-sp.eaking  area  of  Ghana,  where  it

is  the  language  of  trade*  education  and  Christian  religious  observances*

It  has  not  been  possible  to  determine  which'  set  of  names—  that

used  by  speakers-of  Nzema-Jtoyi-Baule  or  that  of  speakers  of  i&an  (Twi-

Fante)  -r was  the  original  one*  It  is. however  clear  that  for  both  group's

of  people  the  week  begins  with  Kwa'sie  (Twi-Fante)  or  Kinzile/feissie

(Nzema-Anyi-Bauie)o 

It  would1 seem  therefore  that  the  apparent  difference

in  the  naming  systems  is  a  result  of  a  'dislocation'  that  took  place  when

each  of  the  two  groups  independently  tried  to  adjust  to  the  European*'

naming  system  for  days  of  the  week*.  It  is  worth  noting  that  on  the

British  calendar  the  week begins  on  Sundays  whije.that  of  the  JFrench

calendar  begins' on  Monday*

The next  question  is  to  determine  the  number  o'f.  language.i  spoken

by  the  Akan peoples,  and  the  relationship  that  exists  between  ihese

languages  arid  their  dialects*

The  question  as  to  how one  determines  whether  a  group  of  dialects

belong  tO' one  language  or. to  different  languages  poses..several  problems*

Che  guch  problem,  which  is  of  particular  relevance  here,  is  the  tact .

that  there  does  not  exist  one  name  for  the  form  of  spoken  language  used

tiy peoples  belonging  to  the  Ak?n  ethnic  group*  The names  Anyi, B<Tule,

Nzeraaj  3ehwi,  Akuapem,  Fante?  e tc  only  refer  to  the  dialects  spoken

by  different  sub/croups  of  the  Aknn  peoples?  and  it  is  quite  clear  that

the  speakers  of  these  dialects  do  not  normally  call  their  language  !Twif .

Stewart  (1966)  subdivides.his  Tano-language  group  into  (a)  the-

Bia  language  group  made  up  of  tvro  languages  -  Anyi-Baule-Cfcako si  and

Nzema-Ahanta,  and  (b)  the  Akan  (Twi-Pgnte)  language• 

In  other  words

the  Akan  peoples  speak  three  languages  -  the  Anyi-Baule-Ghako si  language

with  i ts  dialects  Anyi  and Baule  spoken  in  the  Ivory  Coast,  pnd  CSiakosi

spoken  in  Ghana  and  Togo; 

the  Nzema-Ahanta  language  with  i ts  dialects

Nzema  spoken  in  the  Ivory  Coast  and  in  Ghana,  and  Ahanta  spoken  in

Ghana;  and  the  Akan  (Twi-Fante)  language  spoken  mainly  in  Ghana whose

majsr  dialects  are  Twi  and  Fante»  The  criterion  he  used  in  arriving

at  these  three  distinct  languages  was  that  of  mutual  intelligibility;

rI  have  attempted  here  to  use  the  languages  defined  in  terms  of  the

mutual  intelligibility  of  i ts  idiolect ss  as  the  basic  unit  of  classi-

fication1  . 

(Mutual  intelligibility  refers  to  the  ability  of  different

speakers  to  understand  each  other  without  the  help  of  an  interpreter)

as  is  the  ca.se with  speakers  of  Asante?  Akuapem,  Fante  etc°  in  Ghana.)

In  generals  linguists  have  used  the  criterion  of  mutual  i n t e l l i-

gibility  in  determining  whether  a  group  of  dialects  belong  to  one

language  or  not. 

It  is  however  difficult 

to  determine  the  degree  to

which-dialects  should  be  mutually  intelligible  before-they-  can  be  said

to  belong  to  the  same  language*  Some people  have  argued  that  all  Akan

peoples  speak  the  sane  languages  for  they  claim  that  there  is  a  sufficient

degree  of  mutual  intelligibility  "between  the  various  dialects  spoken  by

people  belonging  to  the  Akan  ethnic  group  for  all  of  them  to  be  considered

as  belonging  to  one  language*  Such people  usually  refer  'to  Modern  Chinese

.and point  out  that  not  all  the  dialects  of  Modern  Chinese''are  mutually

intelligible  but  they  are  said  to  belong  to  the  same  language-

One  of  the  problems  that  the  linguist  faces  in  the  African

situation  however  is  that  very  often,  while  dialects  which  are  mutually

intelligible  have  their  own names?  ©•£•  A'sante?  Panto?  Agon a  etc°.  there

usually  does not  exist  one  name  for  the  language  as  a  wholes  as  in  the

case  of  Chinese  for  example*  The  concept  of  a  language  as  a  single

unit  therefore  has  to  be  forged  by  incorporating  all  the  mutually

intelligible  dialects  into  one  language  and  giving  that  lajn.gu.age  a name-

This  is  what  has  happened  with  the  adoption  of  the  neutral  name  'Akan'

for  the  language  spoken  in  Ghana whose  dialects  include  Asantes  Fante.

iky em ?  Agona,  Gonrua,  Brong  and Akuapem«

In  t h is  discussion  only  the  criterion  of  mutual  intelligibility

will  be  used  in  determining  whether  the  dialects  spoken  by  the  Akan-. •

peoples  belong  to  one  or  more  languages"  as.stated  earlier9  Stewart's

analysis9  based  on  the  criterion  of  mutual  intelligibility.,.  assigns  .....

these  dialects  to  three  languages*  The present  writer's  own  experience,

however,  has  shown  that  there  is  a  high  degree  of  mutual  intelligibility

between  Anyi  and Kzema,  and  between  Baule  and Nzema.  since  even  with

10

her  rather  inadequate  command  of  Nzema,  and  without  any  previous  expe-

rience  of  Anyi  and  Baule»  she  was  able  to  communicate  satisfactorily,

over  fairly  long periods,  with  speakers  of  Anyi  and  Baulp*  Speakers  of

Sehwi  and  Aowin  from  Ghana  who  have  not  had  -any  previous  contact  with

Anyi  speakers  have  also  been  surprised  at  the  extent  to  which  they  can

communicate  with  Anyi  speakers  in  the  Ivory  Coast  without  the  help  of

an  interpreter-

The  conclusion  that  may  be  drawn  from  this  is  that  we have  in

Stewart'  s  Bia  language  group  only  one  language*  which, may  be named

Nzema-Anyi-Baules  whose  dialects  are  Anyi?  Baule*  Chako si » Nzema»  Atlanta*

Sanwi  (Afema)  Aowin,  and  Sehwi,  all  of  ?vrhioh  are  more  or  less  mutually

intelligible» 

It  must  be  pointed  out,  however*  that  Chakosi*  which  ..

bears  closest  resemblance  to  Baule,  is  spoken  in  the  north-eastern.part

of  Ghana  and  spreads  over  the  border  into  Togo,,  aid  with  the  geographical

separation  of  the  two  dialects?  it  is  very  likely  that  Baule  and  Chako si

are  no  longer  mutually  intelligible > and  that  Chako si  is  now  a  separate

language»

The  next  question  is  whether  there  is  a  sufficient  degree  of  mutual

i n t e l l i g i b i l i ty  between  the  dialects  of  the  Nzema-Anyi-Baule  languages

and  those  of  the  Akan  (Twi-Fante)  language  for  all  of  them  to  be  grouped

together  as  one  language-

Speakers  of  Nzeina,  Ahantas  Aowin  and  gehwi  l i v i ng  In  Ghana  do

•understand  various  dialects  of  Akan  (?7/i~1T1ante)  -n.d  t h is  has  been  used

as  an.  armament  for  saying  that  a ll  these  people  speak  the  same  language*

The  mutual  i n t e l l i g i b i l i ty  between  Nzema5  Ah ant a 9  Aowin  and  Sehwi  on  the

one  hand  and  the  Akan  dialects  on  the  other  is  however  more  in  one

direction  than  in  the  other.  Speakers  of  the  Akan  dialects  who  have"

not  had  any  previous  experi<Ece  of  Nsema?  Atlanta?  Aowin  and  Sehwi  do  not

understand  these  dialect  g.  The  reason  for  t h is  state  of  affairs  is  not

difficult 

to  find. 

It  happens  that  in  the  areas  where  Nzema?  Ahanta,

Aowin  and  Sehwi  are  spoken,  the  Akan  dialects?  Fante  and  Asante»  have

prestige  status  and  are  used  in  schools?  in  church  and  for  trade?  so

that  speakers  of  Nzema?  Ahantas  Aowin  and  3ehwi  have  been  exposed  to

Fante  and  Asante  in  a  way  in  which  speakers  of  Fante  and  Asante  have

not  been  exposed  to  Nzema»  Ahanta?  Aowin  and  Sehwi« 

It 

is  a  fact 

that

speakers  of  Nzema  who  are  domiciled  in  the  Ivory  Coast  and  who  have

therefore  not  had  much  -exposure'to  Fante  and  Asante  cannot  normally

communicate  with  Fafite  and  Asante  speakers  without 

the  help  of  an

i n t e r p r e t e r.  The  degree  of  mutual  i n t e l l i g i b i l i ty  between  the  dialects

of  the  TJzema-Ahyi-Baule  language  and  those  of  the 'Ak.an 'language  is

therefore  not  considered  high  enough  for  them  to  be  grouped  together,  as

belonging  to  the  same  language.  However  since  a ll  these  dialects  are,

closely  r e l a t ed  genetically?'  they" have  a  large  number  of  vocabulary

items  in  common?  -and  it 

is  inevitable  that  speakers  of  the  Nzema-Anyi-

Baule  language  and  speakers  of  the  Akan  (Twi-Fqnte)  language?  can 

..

12

usually  identify  in  the  dialects  of  each  other's  language  a  large-number

of  words  that  occur  in  their  own  dialects*

The  conclusion  then  must  be  that  the  /ik-an  peoples  speak  two  lan-

guages"  These  axe

-j)  The  jxIzema-Anyi-Baule  Ian mage  whose  dialects  are  Anyi9
Baule,  Ghakosi?  Nzema,  Aharrta,  Sqfiwi  (Aferaa)  Aowin  and
Sehwio 
It  has  already  been  pointed  out  that  Chakosi
may  very  well  be  a  separate  language  by  now  since  it 
geographically  cut  off  from  all  the  other  dialects  of  the
language?  and  is  surrounded  by  languages  that  .are  not
genetically  related  to  it»  However  it  has  not  been
possible  to  verify  this*

is

2)  The  Akan  language?  whose  dialects  are  Asante*  Akyem,

Bron/Abron,  Wassa,  Kwahu,  Fante?  Agona5  Gomua and
Akuapem.

There  are  regular  phonological  (sound)  correspondences  that  give

further  evidence  for  separating  these  two  languages-'  The  correspon-

dences  and  the  examples  that  i l l u s t r a te  them  are  listed  under  1,  2r 3»

4»  and  7  in  the  Appendix*

They  are:

1*  C o r r e sPo n dence  between  Nzema~Anyi-Baule  / h/  and  Akan  / k /;

In  intervocalic  position  ( i . e.  between  vowels)  / h/  in  Nzema-Anyi-

Baule  corresponds  to  / k/  in  Akan,  as  illustrated  by  the  words  for 

'heart'

and  'husband'.  The  Nzema~Anyi-Baule  language  exhibits  a  phonological

process  known  as  Consonant  Mutation  whereby  some  consonants  change  their

form  in  certain  specific  environments,  as  illustrated  by  the  following

examples:

gc 

h e 's  gon

ho  didn't

I ' ve  eaten

Nzema: 

ko 

yeho

Akan  : 

ko, 

cko

year) go

o arjko

di

di

mill

madi

In  these  examples  Nzema  / h/  corresponds  to  Akan  / k/  in  i n t e r v o c a l ic

position,  / g/  in  Nzema  corresponds  to  / k/  in  Akan  after  a  nasal  consonant,

and  Nzema  / I/  corresponds  to  Akan  / d/  in  i n t e r v o c a l ic  position-

Consonant  Mutation  is  one  important  phonological  c h a r a c t e r i s t ic  of

the  Nzema~Anyi~Baule  language  that  distinguishes  it  from  the  Akan  language.

2«  Correspondence  between  Nzema~Anyi"Eaule  voiced  consonants

and  Akan  voiceless  consonants

Voiceless  consonants  in  the  Nzema~Anyi~Baule  language  become  voiced

in  the  environment  of  a  preceding  nasal  consonant,  so  that  in  the  examples

l i s t ed  in  the  Appendix,  the  voiced  consonants  / z /,  / d /,  / g/  in  the  words

for 

'water1, 

'long'  and  'groundnuts'  in  the  Nzema~Anyi~Baule  language

correspond  to  the  voiceless  consonants  / s /,  / t/  or  / t s /,  / k/  in  Akan«

It  may  be  mentioned  here  that  in  some  subdialects  of  Brong  and

Wassa,  both  Akan  d i a l e c ts  that  are  in  close  geographical  proximity  to

the  Nzema""Anyi~Baule  language,  voiceless  consonants  become  voiced  when

preceded  by  a  nasal  consonant  (Dolphyne,  19^9)•  These  d i a l e c ts  do  not.

however  sh-re  the  more  complex  Consonant  Mutation  system  of  Nzema""Anyi""

Baule-

3«  Correspondence  between  the  presence  of  medial  consonants  in

Nzema~Atiyi ""Baule  and  their  absence  in  Akan

Where  a  medial  consonant  / k/  or  / t J/  occurs  in  the  Nzema~Anyi-

Baule  language  in  the  words  for 

'to  open1  and  ' to  s p l i t'  (Appendix  3)

the  Akan  words  do  not  h»ve  any  medial  consonant'  This  correspondence

is  very  common,  and  is  further  i l l u s t r a t ed  by  the  examples  in  Appendix

tx and  7.

These  phonological  correspondences  also  make  it  possible  for  the

dialects  of  these  languages  to  be  grouped  in  a  way  that  reflects  the

stages  through  which  the  dialects  broke  away  from  each  other.  -For

example,  among  the  dialects  of  the  Nzema~Anyi-Baule  language,  the

correspondences  l i s t ed  under  a,  6  and  8  in  the  Appendix  give  evidence

for  suggesting  that  the  f i r st  s p l it  was  between  Nzema  and  Ahanta  on  the

one  hand  and  Anyi,  Baule,  Aowin  and  Sehwi  on  the  other,  while  those

l i s t ed  under  5  smd  7  separate  Anyi  and  Baule  from  Aowin  and  Sehwi•

What  is  being  suggested  here  is  that  although  there  are  clearly

three  groups  of  dialects  in  the  Nzema-Anyi-Baule  language,  which  are

1)  Nzema  and  Ah ant a,  (2)  Anyi  and  Baule,  and(3)  Aowin  and  Sehwi,  Anyi

and  Baule  are  more  like  Aowin  and  Sehwi,  than  Nzema  and  Ahanta*

The  following  diagram  sums  up  the  grouping  of  the  dialects  of  the

two  languages  in  terms  of  the  stages  through  which  they  broke  away  from

each  other-  The  straight  lines  link  langu-ges  and  the  broken  l i n es

link  dialects-

'Prpto—Jano'

N z ema*~Anyi ~B a ul e

Akarx

Nzema'  Atlanta

Brong,

Asante,  Akuapem,  F a n te

Anyi,  Baule,  Aowin,  Sehwi

\

\

\

Aowin,  Sehwi

Anyi,  Baule,  Chakosi

Asante 

Akuapem,  Fante

/ 

\

' 

/ 

Akuapera 

\
\
Fante

P h o n o l o g i c al  correspondences  between  l a n g u a g es  and  d i a l e c ts  of  a

language  often  i l l u s t r a te  some  linguistic  processes  that  reflect  h i s t o-

r i c al  changes  that  have  taken  place  in  the  language  (Dolphyne,  1976).

One  such  example  isthe  presence  of  a  medial  consonant  in  the  Nzema~Anyi~

Baule  language  and  i ts  absence  in  Akan  (Appendix  3,  1\,  7)- 

It  can  be

said  that  the  loss  of  a  consonant  over  a  period  of  time  is  a  more  likely

occurrence  than  the  introduction  of  a  new consonant  where  none  existed,

so  that  the  Akan  forms  may  be  said  to  be  a  result  of  the  loss'of  the  medial

consonants  /k,  tJ, 

l , r/  that  occur  in.  thu  Nzema~Anyi"Baule  language. 

In

other  words,  the  Akan  forms  are  a  later  de\relopment  from  the  Nzema~Anyi~

Baule  forms.

These  same  sets  ofexamples  (Appendix  3,  &,  7)  also  i l l u s t r a te  •

historical  changes'that  have  taken  place  among  the  dialects  of  the

Nzema-Anyi-Baule  language- 

In  the  examples  l i s t ed  under  3  and..//.in  the 

• .

16

Appendix,  where/tJ/  regularly  occurs  in  medial  position  in  Aowin  and

Sehwi,  / k/  occurs  in  Ahanta,  and  in  Nzema,  Anyi  -?.nd Baule  both  / k/  and

/ t J/  occur  in  this  position*

The  replacement  of  / k/  by  / t //  before  a  front  vowel,  (e*g*  i,  e,

s)  is  a  common phonological  process  in  many  languages^  The  phonological

process  involved  is  known  as  palatalization,  whereby  a / k/  sound  which

is  f  more  'back'  articulation  than  /t«f/  becomes  progressively  articulated

on  the  hard  palate  under  the  influence  of  the  following  front  (or  palatal)

vowel*  This  is  a  well"attested  sound  change,  and  where,  as  in  the  examples

under  discussion,  one  dialect  of  a  language  has  / k/  before  front  vowels

and  another  has  / t J/  in  that  position  it  can  be  said  that  the  dialect

with  the  / k/  sound  has  been  more  conservative  than  the  other  one,  in  so

far  as  it  has  retained  an  older  pronunciation* 

In  other  words  as  far

as  these  examples  are  concerned,  Ahanta  has  been  the  most  conservative

of  the  dialects  of  the  Nzema~Anyi-Baule  language,  while  Aowin  and  Sehwi

have  been  the  most  innovating-

In  the  examples  listed  under  h. in  the  Appendix  /k/  in  Nzema and

Ahanta  corresponds  to  / I/  in  Anyi,  Baule  and  Aowin,  to  / I/  or  / r/  in

Sehwi,  to  /w/  in  Fante  and  Akuapem,  while  in  Asante  a  second  consonant

is  absent.

The phonological  process  that  accounts  for  these  various  pronun-

ciations  may be  stated  in  terms  of  a  'weakening'  in  articulation,  whereby

a  relatively  strongly  articulated  sound  / k/  is  replaced  by  a  less

strongly  articulated  sound  /I/, 

/ r/  or  /w/  and  eventually  gets  deleted,

17

(Boadi,  197/1.).  Four  groups  of  dialects  c/x  therefore  be  identified;

Nzema  and  Ahanta,;  Anyi,  Baule,  Aowin.  and  Sehwi,  Akuapem  and  Fante,

and  finally  As-ante-  These  dialects  h-^ve  been  l i s t ed  here  in  an  order

which  r e f l e c ts  the  various  stages  of  development  of  t h is  particular

sound  change,  from  the  most'conservative1  dialects,  Nzema  and  Ahanta,

to  the  most  'recent'  dialect.;  Asante*

The  l a st  set  of  examples  in  the  Appendix  shows  a  three—way

grouping  of  the  dialects  of  the  two  languages:

1*  Nzema  and  Ahanta  which  have  a  full  consonant—vowel *

nominal  suffix,

2-  Anyi,  Baule,  Aowin,  Sehwi  and  Asante  which  have  a

reduced  suffix  which  is  usually  a  vowel,

3«  Akuapem  and  Fante  which  do  not  have  a  nominal  suffix*

The  deletion  on  / I/  or  / r/  (the  consonants  that  regularly  occur

in  the  Nzema  and  Ahanta  suffixes)  is  a  common  feature  in  the  pronun-

ciation  of  a  number  of  languages- 

In  the  Ga language  for  example,  the

full  form  'nuu  I s'  (the  man)  is  usually  pronounced  'nuus1,  and  in  Akan,

the  Fante  and  Akuapem  forms  'ysreko'  (we  are  going)  and  'borode'

(plantain)  are  pronounced  'yssko1  and  'boodes'  in  Asante- 

In  other

words  the  suffixed  forms  of  Anyi,  Baule,  Aowin,  Sehwi  and  Asante  in

which  the  / I/  or  / r/  of  the  Nzema  and  Ahanta  suffix  is  absent  are  a

l a t er  development  from  the  Nzema  and  Ahanta  forms,  while  the  Akuapem

and  Fante  examples  which  have  no  suffix  at  a l l,  are  even  more  recent.

18

On  the  whole  it  may  be  said  that  the  Nzema~Anyi-Baule  language

has  been  more  conservative  than  the  Akan  language,  in  so  far  as  i ts

d i a l e c ts  have  not  undergone  as  many  sound  changes  as  those  of 

the

Akan  language,  and  that  the  Western  Akan  peoples  speak  a  r e l a t i v e ly

'older'  language  than  the  Eastern  Akan*  The  'age'  of  the  dialects  of

the  two  languages  is  not  however  a  simple  West  to  East  arrangement,

as  can  be  seen  below  where  the  d i a l e c ts  of  each  language  are  arranged

from  l e ft 

to  r i g ht  s t a r t i ng  with  those  that  have  undergone  the  l e a st

change:

N zema-Anyi ~Baule

Nzema 
and  Ahanta 

_  Anyi  and 

Baule 

__  Aowin  and
""  Sehwi

Akan

Brong/Abron 
and  Wassa 

_  As ante  and 

Akyem

-  Akuapem 

-  Fante

It  will  be  seen  from  t h is  arrangement  that  the 

' o l d e s t'  (Nzema

and  Ahanta)  and  the  'newest'  (Fante)  dialects  of  the  Tano  languages

are  geographically  located  in  the  middle  of  the  language  group.

There  have  been  various  attempts  at  determining  the  migratory

movements  of  various  subgroups  of  the  Akan  peoples,  and  some  of  the

evidence  used  has  been  linguistic* 

It  must  be  pointed  out  that  the

'age'  of  a  language,  in  terms  of  how  much  or  how  l i t t le  change  it  has

undergone,  does  not  necessarily  r e f l e ct 

the  age  of  the  people  who  speak

i t. 

It  IIBS  been  observed  however,  that  when  an  area  in  which  a  language

19

is  spoken  is  characterised  by  a  diversity  of  dialects,  that  area  has

been  settled  much  longer  than  an  area  th-^t  is  relatively  unifornl  in  the

form, of • inngu-^ge  used.  For  example  the  diversity  of  dialects  spoken  in

the  Br'ong  area  of  Gh^na,  as  compared  wjlth  the; relative  uniformity  of  the

Asante  dialect,  even  though  it  is  spoken  in  a  much  wider  area,  indicates

that  the  Brong  area  has  been  settled  ljonger  than  the  Asante

It  is  possible  that  as  more  detailed  studies  of  the  various

dialects  of  the  Volta-Bandama  languages  get  completed,  a  clearer

picture  will  emerge  of  the  settlements  and/or  migrations  of  the

speakers  of  these  languages*  For,  the  moment  however,  it  may be  said

that  the  West  to  East  or  East  to  West-spread  of  the  dialects  spoken  by

the  Akan  peoples  seems  to  indicate  that  the  migratory  movements  of  groups

of  present-day  Akan  peoples  were  mainly  in  a  West  to  East  and/or  East  to

West  direction  as  indeed  claimed  in  the  oral  traditions  of.  some  of  them

(Boahen  197ZL).

!-•  Correspondence  between  h:k

20

i sh

heart
husband

Nzema

ahuli
ehu

Ahanta

ahuluni_
chule

Anyi

Baule

ahuli

Aowin

ahuni_
ehua

Seh?4_

ahuni_
ehue

Asante 

Ak uap em /?ant e

akuna
ekunu

akuna
okunu/ikun

2.  Correspondence  between  voiced  and voiceless  consonants  z : s, d :t

wat cr
long

nzule
tindili

nzuro

nzue
tindi

nzue
ti_ndi_

nzue
tindii

nzwie
tindi

groundnut s

rjgati

Qgati

_

_

3»  Coirespondence  between  presence  or  absence  of  medial  consonant

nsuo 
ti  ntini

nkatie

nsu

tsintsin

si

to  open
to  split

buke
kpat.fi

buko
puko

buke
k p a t / i/  kpaii
kpakx_ 
""

but/e
pat Ji

but/e
pat/i

bue

bue/bwei

Conespordence  between  k:  l / r;  w: 4

knoJ
Monday/
Tuesday

kpoke

dzoke

poke

t/oko

kpole 

kpele

dzole

pole

dzule

5.  Coirespondence  between  kp; p

to  split 

kpat/i 

• 

puko

knot 

kpoks 

poko

kpat/J.  kpaki^

kpole 

kpole

patji_

pole

p^ure

dzile

patj^i

pure

dzuo

pax

po

pow

dauw

pow

6.  Correspondence  between  l :n before  nasalised  vowels

21

English 

Nzema

bird
mouth

alum a
elvalo

Ahanta

alum a

Anyi 

Baule

anuma 
nua 

anuma
nua

Aowin

anma
nua.

Sehwi

anma
anua.

Asaute

arvutnaa
anuu

'.  Correspondence  between  k:  t /;  4

to  split 

kpat.fi 

puke

kpat/i_  kpatjj./

to  open
to  spoil

:

buke 
set/i/
seki

b-uke:
sike •

'••  b u ke

;  ' sat/a.

—
—  :

but/e
set/i

but/e
set/1 

•

biie
se_i

8.

Correspondence  between

presence

or  absence

of  nominal

suffixes

A .1 uap em/Panto

anma
ami

pax

bue/bvvei
sei_:

water
breast

nzule
nyefule
e nzule

nzurc
avuno

azuLx

•  \ nz ue
nufre
: 

nz ue:
nufre
anzue

nzue
ny of ue

anzue

nz ue
nyofuni

asuwe

nsuo
nuf uo

asuu

:

nsu 
nufu  •

asu '(wa)

22

1966:

1974

1875

138-j

1974

1976

196 3

1972

1966

Boahen,  A.A«

Chris taller-,  J.G.

Dolphyne,

Gresnbergj  J.G.

Men sail,  A»A-

•Stewarts  J.M«

Westermann,  D.  & Bryan,  M«A»  1952

"The  Origins  of  tfie  Akari'  Ghana,
Notes  and  Queries,  No-  9?  PP°3-iO°

•Who  are  the  Akan?'  Collogue
Ihter-Unlversitaire  Ghana-Cote
d'lvoire^  Lee  Population  Commnnes
de  la  Cote  d'lvoire  et  du  Ghana-
Abidjan,  pp.  66-79=

A Grammar  of 
the  Asante  and  Fanti
language  called  Tshi  (Twi),  Basel-

Dictionary  of  the  Asante  and  Fanti
language  called  Tshi  (Twi),  Basel.

'The  languages  of  the  GhanqrTvory
Coast  Border1  Collogue  I n t e r-
Universitaire  Ghana-Cote  d'lvoire
pp.  438-457«

'The  Brong  (Bono)  dialect  of  Akan •
Brong  Akyempem  ed«  K«  Arhin,  Afram
Pub1Ic a tion s,  Ac cra,  pp.  88-118.

'Dialect  differences  and  h i s t o r i c al
process  in  Akan1  Legon  Journal  of
the  Humanities  Vol«  I I,  PP«15~27»

Lan-raag'es  of  Africa;  The  Hague,  p»8<.

'Foundations  of  Chokosi  Society'
Ghana  Notes  aad^Queries  So«1 2
pp.13-16.

' Ak an  h i n to r r,  son1 e  1 ingu i s t ic
evidence',  Ghana  Notes  and  gftieries
No.9,  pp.  54-57»

Handbook  of  African  Languages?
P a rt  I,  Languages  of  West  Africa,