SUSANNE HACKMACK (BREMEN)*

Case in Selected Grammars of Swahili

Abstract
Comparing the use of case and case-related notions such as nominative, accusative etc. in grammatical
descriptions of Swahili of colonial times with present-day analyses reveals an interesting contrast:
while the majority of early descriptions apply these constructs liberally, modern approaches do not,
stating instead that Swahili does not exhibit any case system at all. One way of explaining this contrast
would simply be to assume that the authors of early descriptions fell into the trap of transferring time-
tested terminology from traditional or Latin grammar to a language with rather different structural
properties. This assumption may certainly hold for a number of grammars, but it does not suffice to
present the whole picture. For this picture to emerge, the theoretical linguistic context of the times
needs to be taken into consideration, i.e. the question of what case could actually stand for in the peri-
od under consideration and how this concept was used in the description of individual languages and
language as a whole. This in turn will show that the above mentioned attribution of a more or less
naïve transfer of terminology by the early authors on Swahili does not do justice to these authors.

1. Introduction
In view of the developments in the field of grammar writing over the past century it is
hardly surprising that the comparison of Swahili grammars from the era of German
colonialism in East Africa with modern descriptions reveals quite a number of differ-
ences. These differences range from the overall organisation of the grammar and its
presentation of the material to the employment of specific constructs, for example terms
such as NP, PP and the like originating from phrase-structure-grammar, which were not
in use during colonial times.
Notwithstanding these differences, modern readers will hardly seem to encounter dif-
ficulties digesting the information conveyed in grammars of colonial times: the majority
of linguistic terms originate in what is commonly called traditional grammar, referring
essentially to

*In: Zimmermann, K. (Ed.) & Kellermeier-Rehbein, B. (Ed.) (2015). Colonialism and Missi-
onary Linguistics. Berlin, Boston: De Gruyter. (51-74)
2 Susanne Hackmack

o major and minor parts of speech such as noun, verb, pronoun etc.,
o secondary parts of speech such as tense, number, case etc. plus their re-
spective values, i.e. past, present etc. or singular, plural etc. or nominative,
accusative etc.,
o grammatical functions such as subject and object.
As concerns terminology, then, Swahili grammars produced during colonial times
have a familiar ring to them as a large number of the constructs applied are still very
much in use in present-day descriptions. However, this supposed familiarity may be
misleading.
A detailed comparison, i.e. one that contrasts the actual application and embodiment of
certain constructs in grammars of colonial times with modern grammars shows interesting
differences concerning the respective interpretations of said constructs with respect to con-
crete language data. One example for this phenomenon is the use of case and case-related
notions, which will be investigated based on the following sample set of grammars:
Büttner, C. G:
1891 Hülfsbüchlein für den ersten Unterricht in der Suaheli-Sprache
Delius, Siegfried und Karl Roehl
1939 Wegweiser in die Suaheli-Sprache
Meinhof, Carl:
1906 Grundzüge einer vergleichenden Grammatik der Bantusprachen
1940 Die Sprache der Suaheli in Deutsch-Ostafrika
Planert, Wilhelm:
1907 Die syntaktischen Verhältnisse des Suaheli
Seidel, August:
1890 Praktische Grammatik der Suaheli-Sprache
1900 Suahili Konversationsgrammatik
Velten, Carl:
1904 Praktische Suaheli-Grammatik nebst Wörterverzeichnis
The authors of this set represent a range of professions that was characteristic for
writers of grammatical work undertaken during colonial times: Seidel, Velten and Plan-
ert were trained as linguists and had no specific affiliation with Christian mission.
Meinhof, Büttner, Roehl and Delius were trained as protestant theologians, the latter
three completing longer stays as missionaries in Africa. Meinhof, whose work was ex-
tremely influential (see below), is difficult to pigeonhole with respect to his scholarly
activities since his œuvre covers a broad spectrum of issues. He is nowadays primarily
associated with work on African linguistics but he also published extensively on other
matters, for example religion, theology and Christian mission on the one hand as well as
questions dealing with the politics of colonisation on the other.
Case in Selected Grammars of Swahili 3

Amongst other things, for example accessibility, this set was chosen because it repre-
sents a cross-section of work on Swahili during colonial times1 and contains work ad-
dressing both learners (Büttner, Delius-Roehl, Meinhof 1940, Seidel, Velten) as well as
a more linguistically informed audience (Meinhof 1906, Planert). Another point of vari-
ation is the relative quality of the grammars: while, say, Delius-Roehl present a rather
consistent and organised account, Seidel 1900 appears to be somewhat confused (and
confusing).

2. Case in Swahili: Modern analyses

In Swahili, a noun phrase such as mpishi 'the/a cook' or mti yule 'that tree' stays invari-
ant irrespective of its relation to either the verb, or another noun, or a preposition.2 The
only instance which may be considered to reflect some kind of overt differentiation
owing to a different relation with respect to the verb is to be found in diverging argu-
ment markers for subject and object in verb morphology as shown in the following
examples:
1. a- li- ku- pig -a
SM TA OM V Mood
CL1(3SG) PAST 2SG hit IND
'He hit you'
2. u- li- m- pig -a
SM TA OM V Mood
2SG PAST CL1(3SG) hit IND
'You hit him'
(SM: subject marker, OM: object marker, TA: tense-aspect marker)
Depending on their grammatical function the agreement markers for subject and ob-
ject alternate: a- vs. m- for third person singular, ku- vs. u- for second person singular.
Krifka (1995) speaks here of "the only trace in Swahili of a case system". Consider-
ing the fact, however, that this divergence appears in a very restricted context only,
namely with noun-phrases in the second person on the one hand and, on the other hand,

1 Delius-Roehl 1939 as well as Meinhof 1940 were published some twenty years after the end of the
German occupation of East Africa. They are written very much in the spirit of the colonial era,
though, as evinced for example in the foreword to the third edition of Delius-Roehl 1939 by
K. Roehl: "So möge denn nun das Büchlein hinausgehen und an seinem Teile mit dazu beitragen,
die Verbindung mit unserem, uns alten Afrikanern unvergeßlichen Deutsch-Ostafrika aufrecht zu
erhalten […]" (Delius-Roehl 1939:IV) 'May this book circulate and help sustain the bond with our
Deutsch-Ostafrika which will remain unforgettable to us old Africans. […]'
2 We will be concerned with verb-governed case exclusively.
4 Susanne Hackmack

exclusively with noun-phrases of class 1 in the third person,3 and further considering
that the precise status of these affixes with respect to their function as an agreement
marker, a pronominal element or both is not conclusively settled, this divergence is
neglectable and the claim that Swahili does not exhibit any overt case marking on nom-
inal constituents is justified.
This claim is found in countless current descriptions, for example in Krifka 1995:
1399, Möhlig & Heine 1999: 71 or in the WALS (World Atlas of Language Structures
Online, Dryer & Haspelmath 2011, see features 28A, 49A and 50A). Creissels
(2005:247) explicitly states that to his knowledge, there is no African language that
exhibits a case system "in which nouns are inflected for case, and where modifiers agree
in case with the noun they modify".
It is thus not surprising that case and case-related notions are hardly used in modern
grammars of Swahili.4

3. Case in Swahili: Analyses from colonial times

Grammars of colonial times present a very different picture. As has been noted by Hen-
nig (2009), these works make ample use of constructs such as nominative, genitive,
accusative and so forth. Analysing her sample of grammars on African languages
(which also includes Meinhof 1906) with respect to case, Hennig (2009:128) notes that
"Obwohl erkannt wird, dass die Einheitenkategorie Kasus offenbar nachrangig ist, wer-
den ihr eigene Kapitel gewidmet. Diese Diskrepanz ist besonders ausgeprägt bei Mein-
hof, der einerseits klar hervorhebt, dass es nie eine Kasusflexion im Bantu gegeben hat
und andererseits dem Kasus ein Kapitel mit Teilkapiteln zum Nominativ, Akkusativ,
Dativ und Genitiv widmet." 'On the one hand, case categories are considered minor
notions, on the other hand they are dealt with extensively. This discrepancy is particu-
larly pronounced in Meinhof, who points out that there never was any case inflection in
Bantu languages, yet devotes a whole chapter to case with subsections on nominative,
accusative, dative and genitive.'
This statement also holds true of the set of grammars listed above: although the pre-
cise way and frequency with which case-related terms are employed may vary, none of
the sample grammars gets by without referring to them, as the following, randomly

3 Like the vast majority of Bantu languages, Swahili exhibits an elaborate gender (or noun-class)
system in which – depending on the underlying classification – up to 15 different noun-classes can
be identified. Nouns of Class 1 (singular) and its corresponding class 2 (plural) prototypically refer
to people.
4 'Modern grammars' refers to work that attempts to give a comprehensive, primarily descriptive
account of a language, working with an inventory of terms as known from Basic Linguistic Theory
(see, for example, Dixon 2010). Approaches operating with dichotomies such as 'deep case' and
'surface case', as, for example, Government & Binding Theory, are not taken into consideration.
Case in Selected Grammars of Swahili 5

chosen examples show. The equivalent of the English term subject is referred to as
either Subjekt or Satzgegenstand, and the equivalent of the English term object is re-
ferred to as either Objekt or Satzziel. The various case-forms are referred to by either
their Latin names, by ordinals (where 1. case = nominative, 2. case = genitive, 3. case =
dative and 4. case = accusative) or by the use of the interrogative pronoun which in
German is fully inflected for case (wer–wessen–wem–wen). These and further features
which will be discussed below make such statements difficult to transfer into English.
The translations are thus to be read cum grano salis.
Eine Deklination der Hauptwörter gibt es nicht im Suaheli. Nominativ, Dativ und
Akkusativ lauten gleich. Ersterer steht immer vor dem Zeitwort, Dativ und Ak-
kusativ nach demselben. Treffen letzere beiden in einem Satze zusammen, so
steht der Dativ vor dem Akkusativ, z.B. mwanamke huyu ana mtoto, diese Frau
hat ein Kind, mwanamume huyu amemwonyesha mtoto maua yale, dieser Mann
hat dem Kinde jene Blumen gezeigt. (Velten 1904:35)
'There is no declension of nouns in Swahili. Nominative, dative and accusative
are identical. Nominative precedes the verb, dative and accusative follow it. If a
sentences contains both dative and accusative, dative precedes accusative, for ex-
ample mwanamke huyu ana mtoto, this woman has a child, mwanamume huyu
amemwonyesha mtoto maua yale, this man has shown that child those flowers.'
Der Wer-Fall (Nominativ) wird gekennzeichnet durch die Voranstellung des
Hauptwortes, das Satzgegenstand ist; und durch seine Wiederaufnahme am An-
fang des Zeitwortes durch die diesem vorangestellte Kennsilbe des Satzgegen-
stands:
Mtoto amechukua kisu
das Kind hat ein Messer genommen
Der Wen-Fall (Akkusativ) wird dadurch gekennzeichnet, dass das Satzziel (Ob-
jekt) hinter das Zeitwort gestellt wird:
Kisu kikali kimekata mtoto
das scharfe Messer hat ein Kind geschnitten
Der Wem-Fall (Dativ) wird dadurch gekennzeichnet, dass das entferntere Satz-
ziel des Wem-Falles vor das nähere des Wen-Falles gestellt wird:
nimempa mtoto mzigo mdogo tu
ich habe dem Kinde nur eine kleine Last gegeben. (Delius-Roehl 1939:50)
'Nominative is characterised by prepending the noun which is subject and refer-
encing it at the beginning of the verb by means of a subject index-marker:
Mtoto amechukua kisu
the child has taken a knife
Accusative is characterised by positioning the object behind the verb:
Kisu kikali kimekata mtoto
6 Susanne Hackmack

the sharp knife has cut a child

Dative is characterised by positioning the indirect object in front of the object:
nimempa mtoto mzigo mdogo tu
I have given the child only a small load.'
Der Nominativ ist der Kasus des Satzsubjekts und des Prädikats, sofern es ein
Substantiv (bzw. Fürwort ist, z.B.
Subjekt: muungu anaweza yote, Gott vermag alles
Prädikat: mimi si dada yako, ich bin nicht Deine Schwester. […]
Manche Verba regieren im Aktivum zwei Akkusative, deren zweiter im prädika-
tiven Verhältnis zum ersten steht. Wendet man Sätze mit derartigen Verben pas-
sivisch, so werden beide Akkusative in den Nominativ verwandelt. (Seidel
1900:254)
'Nominative is the case of the subject and the predicate, if this is a noun or pro-
noun, for example:
subject: muungu anaweza yote, God is capable of anything
predicate: mimi si dada yako, I am not your sister […]
Some verbs govern two accusatives in the active form, the second of which is in
a predicative relation to the first. If such sentences are passivized, both accusa-
tives change into nominatives.'
Die Formen der persönlichen Fürwörter sind sich gleich für den dritten und vier-
ten Fall, und unterscheiden sich von den Formen des ersten Falls nur in der zwei-
ten und dritten Person (der mtu-Klasse). Bei den Hauptwörtern ist es meist nur
aus der Stellung im Satz zu sehen, ob sie im ersten oder im vierten Fall stehen.
(Büttner 1891:17)
'The forms of personal pronouns are identical in dative and accusative and differ
from the forms of nominative only in the second and third person (of the mtu-
class). As concerns nouns, the only way to identify whether they are nominative
or accusative is by their position within the sentence.'
Even Planert, who specifically points out that "Es bedarf keiner besonderen Ausei-
nandersetzung mehr, um das im Indogermanischen geltende Kasussystem als für die
Gesamtheit der Sprachen durchaus unzulänglich zu erweisen." 'There is no need for
discussion anymore to prove that the Indo-Germanic case system is quite inappropriate
for the entirety of languages' (Planert 1907:12) repeatedly uses constructs such as accu-
sative, dative etc. and compound expressions such as nominative/genitive or nomina-
tive/locative in his description of Swahili.
Irrespective of the quality of the above translations, one thing is obvious: compared
to modern grammars the descriptive category case seems to have undergone a concep-
tual shift since its application to Swahili in modern analyses (namely, no application) is
diametrically opposed to the way it was used during colonial times.
Case in Selected Grammars of Swahili 7

This conceptual shift is particularly obvious in statements such as "nominative is char-

acterised by pre-verbal position", i.e. statements that define a case-form on the basis of its
position in the sentence, surely something that no present-day author would do.

4. An attempt at explanation
One possible reason for the way the category case was used in grammars of colonial
times is based on the assumption that the authors applied terminology familiar from
grammars of German (or Latin) more or less naïvely to a language that exhibits quite
different structural features. Amidu (2004:47) calls grammars that operate on this basis
"translational grammars" and describes them as follows: "They are founded on the prin-
ciple that given knowledge of English or French grammar, all one needs to do in writing
a grammar of Kiswahili or other unwritten languages, is to present the material in such
a way that English or French nationals can easily understand it."
A translational grammar in the sense of Amidu would involve the translation of
Swahili expressions into German, the analysis of the German translations with con-
structs from traditional or Latin grammar and the application of said constructs to the
corresponding elements in the Swahili data. In essence, this modus operandi constitutes
(at least in part) what Hennig glosses a "German perspective" (cf. Hennig 2009). If this
procedure involves an uncritical transfer of a formal concept that has no formal corre-
late in Swahili (as, for example, dative), it may well be viewed as naïve or even show-
ing a colonial mentality.
Hennig and Amidu are not the only authors who find fault with such practice.
Welmers has voiced the following criticism:
In conclusion, African language studies have too commonly – and sometimes
still are – characterized by an uncritical and often naïve imposition of classical
grammatical categories and terminology in structures to which they are quite for-
eign. In both description and labelling, it is past time to bring the era of linguistic
imperialism to an end and make a more consistent effort to capture what is really
going on in African languages. (Welmers 1973:382–383)
Welmers' reprimand seems a bit harsh considering the complexity of the task at hand:
the prerequisite for capturing "what is really going on" in any language, i.e. a set of
structural tools that would account for language-specific idiosyncrasies while at the
same time being adequate for use in language comparison, has not been compiled as yet
and constitutes a primary topic in linguistic theory-building to this very day (cf., for
example, the Haspelmath-Newmeyer-discussion in Language 86(3), 2010).
As concerns the sample grammars of Swahili, accusing the authors of linguistic im-
perialism is tempting considering the overall imperialistic setting and frame of mind in
which they worked but overlooks the fact that the majority of these grammars were
8 Susanne Hackmack

intended to be used in teaching and that many authors used terms and constructs that
they considered their potential audiences familiar with, putting up with the shortcom-
ings of this approach quite consciously.
And finally, one important point needs to be clarified before we ascribe "an uncritical
and often naïve imposition of classical grammatical categories and terminology" to the
sample grammars, namely how case and case related notions were generally interpreted
and used at the end of the 19th and the beginning of the 20th century and how the authors
of these grammars may have been influenced by the linguistic theory of the times. The
following sections are devoted to the discussion of this point.

5. Case then and now

In order to pinpoint differences between the use of case and case-concepts during colo-
nial and modern times, a first approach may be to compare standard definitions of case
from the 19th and early 20th century with modern definitions. However, such a compari-
son does not lead very far. As a matter of fact, definitions in standard encyclopaedias
from the 19th and early 20th century and modern terminological dictionaries are not at all
out of accord. As an example, compare the following entries for case from the 19th, 20th
and the 21st century:
Herders Conversations-Lexikon 1857:
Beugungsformen, Fälle, nennt die Grammatik die verschiedenen Formen eines
Nomens, um dadurch die Beziehungen auszudrücken, in welche es im Zusam-
menhang der Rede als Satztheil, gekommen ist.
'In grammar, inflectional forms or cases are the various forms that a noun may
have in order to indicate the relations that it may enter into in connected speech'
Pierer's Universallexikon 1857:
Beugefall od. die Veränderungen, welche am Ende des Stammes od. am Stamm-
vocal eines Nomens vorgenommen werden können, damit die verschiedenen
Verhältnisse der Objecte zu dem Subjecte in dem Satze angezeigt werden kön-
nen.
'Inflection or change that may occur at the end of the stem or the stem vowel in
order to indicate the various relations of the objects to the subject in a sentence.'
Case in Selected Grammars of Swahili 9

Brockhaus' kleines Konversationslexikon 1911:

Kasus (lat. Casus), Beugefälle, in der Grammatik die verschiedenen Formen,
welche ein Nomen oder Pronomen zum Ausdruck der verschiedenen Beziehun-
gen im Satz (Subjekt, Objekt etc.) annimmt.
'In grammar, cases are the various forms that a noun or pronoun takes in order to
express the various relations in a sentence (subject, object)'
The Concise Oxford Dictionary of Linguistics 22007:
Inflectional category, basically of nouns, which typically marks their role in rela-
tion to other parts of the sentence. E.g. in Latin vidi puellam 'I saw a girl' puellam
'(a) girl' has the ending of the *accusative case (puella -m) and this marks it as
the object of the verb vidi. (Matthews 2007 s.v. case)
Judging solely from these definitions, the difference in the application of case and
case-related notions with respect to Swahili seems none the clearer. The entries are all
similar and name two parameters related to case, one of which refers to its more formal,
morphological aspect, i.e. inflection, the other referring to the potential function of
case-forms, i.e. marking relations in a sentence. These relations are specified in terms of
grammatical functions, i.e. subject and object.
Explications such as the ones above are problematic, though, in two respects:
1. it is in their very nature and owed to their need for brevity not to be concerned
with the often complex and controversial discussion of the definiendum within
theoretical linguistics,
2. they are based on a number of defining terms which themselves need to be clar-
ified in order to fully grasp the entry at hand (here: grammatical functions).
A deeper understanding of the discrepancy between grammars of Swahili in colonial
and modern times will thus have to come about by a more thorough study of the discus-
sion of case and notions such as subject and object during the time in question.
To all intents and purposes comparable to modern linguistics, case and case-related
concepts were lively debated issues at the end of the 19th and the beginning of the 20th
century within German theoretical linguistics and had been so for some time before, too
(cf. Hübschmann 1875:74-130). The following list shows that the topics addressed were
manifold and often have a decidedly modern ring to them. As a matter of fact, many of
these issues have not yet been conclusively settled and seemingly similar questions
continue to be discussed up to the present day (see, for example, Cienki 1995 and Wil-
lems 1997).
Controversies touched topics such as
o the differentiation of so-called logical and localist cases (see below) and the
question whether the latter can be assumed to be primary notions, underlying
all cases, or not (see, for example, Holzweissig 1877, Marty 1910),
10 Susanne Hackmack

o the question of whether specific case-forms may be allocated a stable basic

meaning (Grundbedeutung or Grundbegriff) and if so, what this meaning is and
how it can be described,
o the discussion of the various uses (Gebrauchsweisen) of a specific case-form
(such as dativus commodi, dativus incommodi, dativus possessivus etc.) (see,
for example, Delbrück 1907:225ff),
o methodological questions such as whether case and case-forms ought to be ap-
proached semasiologically or onomasiologically, i.e. whether the starting-point
for analysis ought to be the set of forms that a given language exhibits or the
set of relations that these forms may potentially express. Brugmann (1904:373)
comments on the problems of the wide-spread onomasiological approach as
follows: "Die Gruppierung und Benennung der Kasusbildungen geschieht her-
kömmlicherweise nach der Bedeutung, nicht nach der Form, was zur Folge hat,
dass zumteil nicht nur formantisch Verschiedenes unter eine Benennung fällt,
[…] sondern auch formantisch Identisches oder doch im formantischen Wort-
teil etymologisch Verwandtes voneinander getrennt wird […]".
'Grouping and naming of case-formation is conventionally done on the basis of
meaning rather than form which means that in some instances, formally distinct
entities are subsumed under one designation […] while in other instances, for-
mally identical or etymologically related entities are being separated […]'
A thorough review of these discussions and the question of the degree to which they
may have foreshadowed modern developments would be a very valuable endeavor but
lies beyond the scope of this paper. In any case, the list shows that there were many
open questions and varying opinions in connection with case. This means that in order
to make sense of the way that an individual author of colonial times interpreted and
applied case-notions or grammatical concepts in general, it is advisable to study the
theoretical 'affiliation' of said author, i.e. the question by which of the proposed ap-
proaches or by which authors he or she may have been influenced. This task is by no
means limited to authors of the past but also holds true concerning present-day work.

6. Linguistic Influences
As concerns the authors of the set of grammars under consideration, the central figure is
surely Carl Meinhof, whose work can be considered to have been the cornerstone for
the development of African studies in Germany. The following, slightly longer excerpt
from Meinhof's obituary in African Studies underlines his massive impact on the field
and its students:
It would be extremely difficult to overestimate the value of the work done by
Meinhof in the field of African linguistics. In its combination of quantity and
Case in Selected Grammars of Swahili 11

quality, of range and depth, it holds a unique place. In the course of his fortunate-
ly long career – though the life of a man such as he must, to his friends and ad-
mirers, always seem too short – this great scholar engaged in the most numerous
and varied activities for the promotion of the science which he had made his
life's work, and with which his name must always remain prominently and grate-
fully associated. In his extensive researches and publications – kept up, we know,
until the very eve of the latest world war, and continued, we may feel sure, dur-
ing that war as it was during the previous one – he penetrated into every corner
of the African linguistic field, and made, besides, more than one contribution of
value to African studies other than linguistic, and to linguistic studies other than
African. In his work as a teacher and guide of others interested in African lan-
guages and allied subjects, he gathered round him, both to his own school in
Hamburg, and in places geographically far removed from that Mecca of the Afri-
can linguist, an ever-growing band of pupils and collaborators, whom he in-
spired, stimulated, and in every way aided in the quest for knowledge of the Af-
rican and his tongues; and, with Meinhof at its head, and as its chief forum the
journal which he founded as the Zeitschrift für Kolonialsprachen (renamed the
Zeitschrift für Eingeborenen-Sprachen when Germany had lost her colonies), the
Hamburg school took and long retained the lead in African linguistic science,
and, to say the very least, remained unquestionably in the very front rank when
other schools of African languages became active elsewhere. (Lestrade 1946:73)
With the exception of August Seidel, Meinhof's close association with the other au-
thors of the above set (or, rather, their close association with Meinof's work) can be
safely established. Büttner was a "good friend" of Meinhof, Meinhof's and Velten's
tenure at the Seminar für Orientalische Sprachen in Berlin overlapped, Roehl and Deli-
us were students of his and mention him both in their respective forewords and Planert
explicitly thanks Meinhof and names him as influential for his work (Pugach 2011:73,
Delius-Roehl 1939:I & IV, Planert 1907:VI, and the respective entries in Jungraithmair
& Möhlig 1983).
On the basis his of pivotal position in the newly developing field of African studies
and the assumption that his viewpoints and methods have at least partially rubbed off on
others, the question of which linguistic thought Meinhof himself was influenced by is of
central interest.
Establishing to what precise degree specific work or specific authors may or may not
have inserted influence on others can prove to be a difficult task, growing all the more
complicated the more time has elapsed. Koerner, who devotes quite some work on this
question (see as an example Koerner 1989), mentions various ways in which the exer-
tion of linguistic influence may be detected. Of these, "the most important evidence in
favour of a claim of influence may result from direct references by an author to the
work of others." (Koerner 1989:41).
12 Susanne Hackmack

In the case of Meinhof, there is one author who undisputedly exerted a strong influence
and thus deserves special attention, namely Wilhelm Wundt and his school of thought.
Wundt (1832–1920) is probably best known as one of the founders of Völkerpsy-
chologie ( 'psychology of peoples'), which is described as follows:
Völkerpsychologie heißt seit Mitte des 19. Jahrhunderts die Wissenschaft vom
Volksgeist, von den Elementen und Gesetzen des Seelisch-Geistigen im Völker-
leben. Die V. untersucht jene psychischen Vorgänge, die in ihrer Entstehung u.
Entwicklung an menschl. Gemeinschaften gebunden sind. Ihre Hauptuntersu-
chungsgebiete sind die Erscheinungen des objektiven Geistes, vor allem Sprache,
Kunst, Mythus und Religion, Sitte, Gesellschaft, Recht, Kultur überhaupt.
(Schischkoff 1957:631)
'From the middle of the 19th century onwards, psychology of peoples is the sci-
ence of ethnic spirit, of psychological elements and rules that are manifest in eth-
nic communities. Psychology of peoples studies those phenomena whose emer-
gence and development are bound to human society. The primary areas of
research are aspects of the objective spirit, such as language, art, myths and reli-
gion, moral conventions, society, law, or, generally speaking, culture'5
First and foremost a psychologist, then, Wundt devoted the first two books of his
volumes on Völkerpsychologie to language and it is thus not surprising that he was an
integral part of the discussion within general linguistics, too. Accordingly, Wundt was
studied widely and read – amongst others – by Meinhof. This is not only evident from
the fact that Meinhof cites Wundt repeatedly, but can also be deduced from other
sources, which underline the strong bond between Wundt and Meinhof. Wundt wrote
the following praise of Meinhof:
Prof. Meinhof hat sich durch seine Lautlehre der Bantusprachen sowie durch ver-
schiedene andere Beiträge zur afrikanistischen Völkerkunde ausgezeichnet. Er
kann als ein hervorragender Vertreter jener aus der deutschen Mission [...] her-
vorgegangenen Männer betrachtet werden, die sich um die Erforschung der
Sprachen und Kulturen vornehmlich der dem Umkreis der deutschen Kolonial-
gebiete angehörigen Völker in hohem Maße verdient gemacht [...] haben. (Wundt
1909, quoted from Brauner 1995:59)
'Prof Meinhof has distinguished himself with his study on the phonology of Ban-
tu languages and various other contributions to African ethnology. He can be re-

5 Please note that the term 'psychology of peoples' and the following translation can only be consid-
ered an approximation as constructs such as Volk, Völkerpsychologie, Volksgeist, Erscheinungen
des objektiven Geistes etc. are anything but clearly defined notions. 'Völkerpsychologie' itself has
been translated into English in a number of ways, for example as 'folk psychology' or 'ethnic psy-
chology' (cf. Blumenthal 1975:1086, fn. 1)
Case in Selected Grammars of Swahili 13

garded as an outstanding representative of the German mission and those men

that have rendered excellent service in researching the languages and peoples in
and about the German colonial territories'
Meinhof in turn states that it was the study of Wundt's Völkerpsychologie that ena-
bled him to structure and organise all the many particulars that he encountered in his
work over the years (Meinhof 1917, quoted from Brauner 1995:64) and stresses the
importance of Wundt' s work for linguistics in general and African studies in particular
(Meinhof 1920:241ff).
As concerns case, Wundt (1904:71ff) held the view that the development of case in a
language follows a fixed pattern of stages. The first (or, in Wundt's words, lowest) stage
is evident in languages "bei welchen nicht nur Wortunterschiede, sondern auch andere
Ausdrucksmittel der Kasusbeziehung bloß in schwachen Spuren vorkommen", i.e. in
which the means to express case-relations, such as modification of the noun or other
means, are rare. Wundt names various African languages as well as the languages spo-
ken on Papua-New Guinea as examples. The second stage is more or less the opposite
of the first and is found in languages that exhibit an excessive degree of nominal modi-
fication, in which the type of relation expressed also comprises 'concrete', i.e. spatial
and temporal relations. According to Wundt, this type is manifest in languages spoken
in northern America, Oceania and Australia, the "highest degree of excessiveness",
though, is found in Basque and the proto-Altaic and Caucasian languages. The third
stage, finally, is represented by Indo-European and Semitic languages in which case-
formation (on the noun) is restricted to mark but a few, as Wundt calls them Grundver-
hältnisse ('basic relations') while other relations (again primarily spatial) are expressed
by the use of adpositions or particles which exclusively serve this purpose.
Wundt considers as basic those relations that can be mapped onto grammatical func-
tions such as subject and object and these he calls cases of "innere Determination"
( 'internal assignation') whilst other, more "remote" relations are called cases of
"äußere Determination" ( 'external assignation'). His whole theory on case hinges on
this classification.
However, Delbrück (1901:129) points out that Wundt's categorisation of cases leads
to precisely the same classes as in the time-honored division of cases into "grammati-
cal/logical" cases such as nominative, accusative, dative etc. and "local" cases such as
ablative, locative, essive etc. This is surprising, for Wundt objected to this division.
Delbrück, who devotes a whole book to Wundt's linguistic theory, was not the only one
who found fault with his approach. Another vocal critic was Marty (Marty 1910), who
also notices that Wundt uses elements of a theory he disapproves of to substantiate his
own, very similar classification ("Etwas, was er der von ihm bekämpften Theorie zum
Vorwurfe macht, trägt er also auch wieder unbedenklich als Bestandteil seiner eigenen
vor." Marty 1910:7).
14 Susanne Hackmack

Surely one of the biggest problems in Wundt, however, is the fact that the terminolo-
gy he used was anything but clearly defined, leading to severe problems in understand-
ing and even contradictory statements. Concepts such as Kasus 'case', Kasusbegriff
'case-concept', Kasusform 'case-form', Kasusverhältnis 'case-relation', Kasusbedeutung
'case-meaning', logisch 'logical', psychologisch 'psychological', grammatisch 'grammati-
cal' etc. were used in an extremely inconsistent manner as evinced by the plethora of
footnotes in Marty (1910) devoted to Wundt's "Ungenauigkeiten" ('imprecisions').
Not least because of this terminological confusion a detailed review of Wundt's work
on case is beyond this paper. The focus thus lies on those assumptions that most clearly
can be seen to have had an impact on Meinhof. As the following two excerpts show,
there are strong parallels in Wundt's and Meinhof's stance on case:
Diese Vorgänge der Verschmelzung und Differenzierung der Kasus unter dem
Einfluß mannigfacher Assoziationsbedingungen durchkreuzen sich nun noch mit
zwei weiteren Erscheinungen von entgegengesetztem Charakter, die aber beide
dahin zusammenwirken, daß die Kasusformen des Nomens überhaupt für die ei-
ner Sprache zur Verfügung stehenden Kasusbegriffe durchaus kein Maß abgeben
können. Erstens kann nämlich die Sprache gewisse Kasus bloß durch die
W o r t s t e l l u n g ausdrücken, ohne daß am Worte selbst irgendwelche Verän-
derungen eintreten, die mit dem Kasusbegriff in Beziehung stehen. Zweitens
können sich besondere, von dem Worte trennbare Partikeln entwickeln, die als
Äquivalente der Kasusformen funktionieren. Solche Partikeln sind die Präpositi-
onen, an deren Stelle in selteneren Fällen auch Postpositionen auftreten. Infolge-
dessen kann eine Sprache an den spezifischen Kasuselementen des Nomens, den
Kasussuffixen oder -präfixen, sehr arm sein und gleichwohl über eine reiche Fül-
le wirklicher Kasusunterscheidungen verfügen. (Wundt 1904:68-69)
' Those processes of conflation and differentiation of case under the influence of
associative relations are interwoven with two further phenomena of converse
character, which both lead to the insight that the case-forms for nouns supplied
by a language are no indication whatsoever of the case-concepts that said lan-
guage may have at its disposal. First of all, case may be expressed by w o r d -
o r d e r alone, without the word itself showing any modification that may be re-
lated to this case-concept. Secondly, special particles may develop which are
separable from the word itself and function as equivalents of case-forms. Such
particles are prepositions, in more rare cases postpositions, too. Thus, a language
may be very impoverished concerning case-specific elements of nouns, i.e. case-
suffixes or case-prefixes, yet avail itself of a rich inventory of true case distin-
guishing means.'
Wenn man von Flexion spricht, denkt man zunächst an die Abwandlung des
Nomens und des Verbum durch Affixe, also an Deklination und Konjugation. Es
Case in Selected Grammars of Swahili 15

unterliegt keinem Zweifel, daß viele, wohl die meisten der heute gesprochenen
Sprachen eine solche Abwandlung des Nomen nicht kennen – mag das Nomen
im Satz diese oder jene Stellung haben, es bleibt vollkommen unverändert. Da-
mit ist natürlich nicht gesagt, daß die gegenseitige Beziehung der Worte im Satz,
die z.B. im Lateinischen durch Kasusendungen ausgedrückt wird, überhaupt
nicht angedeutet werden könnte. Es muß ja in allen Sprachen der Welt erkennbar
sein, ob z.B. die drei Begriffe Mann, schlagen, Knabe bedeuten sollen, daß der
Mann den Knaben schlägt, oder daß der Knabe den Mann schlägt. Ferner muß ja
erkennbar sein, ob die beiden Begriffe Vater, Freund heißen sollen "der Vater
des Freundes" oder "der Freund des Vaters" Also muß die Kasusbeziehung über-
all zum Ausdruck kommen. […] Wundt hat vorgeschlagen, zwischen Kasus der
inneren Determination, die überall zum Ausdruck kommen müssen, und Kasus
der äußeren Determination zu unterscheiden, da die letzteren eine entferntere Be-
ziehung ausdrücken und in vielen Sprachen durch Präpositionen und Postpositio-
nen oder andere Umschreibungen ersetzt werden." (Meinhof 1936:83–84).
'The first thing one associates with inflection is the modification of nouns and
verbs by affixes, i.e. declension and conjugation. There is no doubt, however,
that many, probably most of the languages spoken today do not exhibit any such
modification of the noun, which stays invariant whatever its position in the sen-
tence. This is not to say that the mutual relation of words in a sentence, which in,
say, Latin is expressed via case-endings, cannot be expressed. Every language in
the world needs to be able to indicate whether the three terms man, hit, boy are to
mean that the man hits the boy or the boy hits the man. Further, it needs to be
able to indicate whether the terms father, friend ought to mean 'the father of the
friend' or 'the friend of the father'. […] Wundt suggests differentiating between
cases of internal and cases of external assignation. Cases of internal assignation
must be expressed in any language while cases of external assignation, express-
ing a more remote relation, may in many languages be replaced by prepositions,
postpositions and other means.'
Both Wundt and Meinhof differentiate between case as a conceptual notion (Kasus-
begriff 'case-concept' in Wundt, Kasusbeziehung 'case-relation' in Meinhof) and the
formal means by which this notion can be expressed. As concerns the latter, inflectional
affixes are but one possible means to express the conceptual notion of case – Wundt
specifically names word-order and adpositions as other potential markers and Meinhof
adopts this point with respect to cases of "external assignation".
In these quotes both Wundt's and Meinhof's use of the term case – and, as will be
shown below, case-related notions such as nominative etc. – correlates primarily with
the functional aspect of this concept, which means that 'case' can be equated to 'relation
in a sentence' and is thus dissociated from the formal aspect of inflection. It is on this
16 Susanne Hackmack

basis that Wundt and Meinhof could consider cases "of internal assignation" to be a
universal concept.
This observation is crucial for the discussion at hand, namely the discrepancy in the
application of case and case-related notions then and now as described above: if case
could be used in a way that is more or less synonymous with 'relation in a sentence' and
which disregards the inflectional aspect of this concept, it comes as no surprise that
Meinhof and others had no qualms in applying this term in the description of Swahili
(or any language without case inflection). In modern grammars, on the other hand, this
use of case seems outmoded.
Thus, an answer to the core-question of this paper is arrived at. At the same time, an-
other historiographically interesting question arises: how did the conceptual shift that
case and case-related notions have undergone over time come about? The answer to this
question will necessitate a closer look at what has been termed quite generally 'relations
in a sentence'.

7. Relations in a sentence
Analysing the set of colonial grammars of Swahili with respect to the question which
relations (verb-governed) case-notions such as nominative, accusative etc. marked, the
answer seems to be in accord with the definitions of case presented above: 'relations in a
sentence' maps onto 'grammatical functions', i.e. subject, object, indirect object. This
allocation is done with varying degrees of explicitness. Meinhof (1906: 27–28) directly
correlates nominative and subject, accusative and object. As has been shown above,
Delius-Röhl equate nominative, accusative and dative with Satzgegenstand, Satzziel and
entferntes Satzziel ' topic', ' sentence target' and ' remote sentence target' which – at
the time – corresponded to subject, object and indirect object. Velten does not explicitly
equate case-forms with grammatical functions, but gives examples which show that he
has, in fact, such a correlation at the back of his mind for he describes case-forms in
terms of word-order, i.e. by explicating how the grammatical functions in Swahili are
identified.
This mapping does not seem too remote from modern uses, i.e. the allocation of case-
forms and specific grammatical functions is still being undertaken as the modern definition
above shows where accusative case is said to denote objecthood.
What has changed massively, though, is the elaboration and specification of the set
of grammatical relations, which came about not least because of the work done on other
types of relation in a sentence.
Generally speaking, we can say that modern linguistics avails itself of a vast invento-
ry of specific terms to denote various kinds of relation within a sentence: apart from the
grammatical functions subject, object etc. there are sets of relations such as source,
goal, agent etc., i.e. terms enumerating the semantic roles of the relation denoted by a
Case in Selected Grammars of Swahili 17

specific case-form. Furthermore, there are terms such as given/new or theme/rheme or

topic/focus etc. to specify a constituent's function with respect to information structure
and sentence perspective.
The use of these constructs may not in all cases be consistent, their precise definition
is still a matter of debate and the relative significance they have within a modern theory
of grammar is as yet not settled. Still, they can be allocated a relatively stable core
meaning: any contemporary linguist will have at least a basic understanding when it
comes to expressions such as subject, agent or theme, and – most importantly – will not
confuse a relation such as subject with either agent or theme. If this were not the case,
one of the most widely-spread systems within syntactic typology would lose its concep-
tual basis: the description of sentences with respect to the three parameters S (only ar-
gument of an intransitive verb), A and P (agent-like and patient-like argument of a tran-
sitive verb, cf. Dixon 1972 or Comrie 1978) hinges on an analysis of sentence-relations
which is more sophisticated than just subject and object.
This state of affairs was rather different during colonial times, as the study and sys-
temisation of these constructs as well as the discussion concerning their place in gram-
matical theory took place after the structuralist turn in linguistics and consequently,
grammarians of the 19th and early 20th century did not have them at their disposal in the
same way.
This is not to say that concepts such as, say, agent, were unknown. However, owing
at least partly to the fact that the analysis of sentences for which these constructs are
essential notions was not undertaken in the same way as in today's linguistics, where
various levels of description (morphological, syntactic, semantic and pragmatic) can be
identified, each of which based on a large theoretical apparatus, these constructs were
used in a much less systematic way at the time in question.
In many instances, then, constructs such as subject or object – and accordingly, nom-
inative, accusative, dative – were used to subsume a number of different types of rela-
tion for which modern linguistics offers specific terms: while modern linguistics distin-
guishes relatively consistently between, say, nominative, agent, subject and theme,
linguistics at the turn of the century, for lack of a systematic terminology, sometimes
had no option but to merge these constructs or, more specifically, certain aspects of
their meaning. A term like subject could thus either refer to the formal, inflectional
features of a noun-phrase (case-form, verbal agreement, position with respect to the
verb etc.), or to a semantic relation comparable to agent, or to a pragmatic notion such
as 'that which is being talked about', 'the starting point of the message' – or, in a simple
active-declarative with an agentive verb, where these relations coincide in a constituent,
all of the above.
Actually, these various aspects of subject and, for that matter, nominative, have all
cropped up in the grammars under discussion. The term wer-Fall 'who-case' used by
Delius-Roehl (see above) refers to inflection and addresses the formal features of nomi-
native/subject, as do specifications that draw on word-order and pre-verbal position.
18 Susanne Hackmack

Satzgegenstand 'Topic' on the other hand corresponds with a pragmatic function com-
parable to theme while Meinhof's statement on case (see above) suggests a reading
more in accord with semantic roles.
Linguists at the time were of course aware of a problem, especially in those cases
where the equation of say, subject with agent or subject with topic would not work (cf.,
for example, von der Gabelentz 1901:370). As a matter of fact, the end of the 19th and
the beginning of the 20th century marked the beginning of a more advanced nomencla-
ture and the discussion of distinctions such as 'logical', 'grammatical' and 'psychological'
subject. The following quote from Wegener (1885: 20) is an example of the way such
concepts were discussed (for a more detailed overview see Seuren 1998:120–133; for a
thorough analysis Elffers-van Ketel 1991:163–326).
[…] die Gruppe von Vorstellungen von der eine Aussage gemacht wird, nennen
wir Subject, die Aussage selbst Prädicat. Das Subject ist das intresselose Bekannte,
die Aussage das Intressierende und Neue, allerdings nicht immer findet dies Ver-
hältniss zwischen grammatischen Subject und grammatischen Prädicate statt. Bei
der Betonung: d e i n Vater hat es gesagt, ist das Neue und interessirende das
grammatische Subject, aber logisch das Prädicat. Man darf darum jene Exposition
das logische Subject, das Interessierende und neue dagegen das logische Prädicat
nennen. Allerdings ist dabei der Uebelstand, dass der Ausdruck logisches Subject
ein fester Terminus in der Grammatik schon geworden ist: Man versteht darunter
das handelnde Subject, besonders wenn dies die Form des grammatischen Sub-
jects, den Nominativ, nicht hat, wie in dem Satze: der Baum ist vom Knaben gese-
hen, hier ist logisches Subject vom Knaben. Vorzuziehen ist darum der Deutlich-
keit wegen statt logisches Subject Exposition zu sagen. (Wegener 1885: 20)
' The group of ideas about which some statement is being made is called subject,
the statement itself is called predicate. The subject is the uninteresting given, the
statement made about the subject is interesting and new. However, this relation
does not hold between the grammatical subject and the grammatical predicate in
all cases. In y o u r father said it, i.e. with the stress on your, the grammatical
subject is interesting and new but logically the predicate. Such exposition may
thus be called the logical subject, that what is interesting and new the logical
predicate. Unfortunately, however, logical subject has become a fixed term with-
in grammar where it is understood to denote the acting subject, especially in
those cases in which it does not exhibit the formal feature of the grammatical
subject, i.e. nominative, as for example in the tree has been seen by the boy,
where by the boy is the logical subject. In the interest of clarity, it would be ad-
visable to use the term exposition instead of logical subject.'
One question worth pursuing would be how well such 19th century concepts compare
with modern concepts such as subject, agent and theme or distinctions such as given/
Case in Selected Grammars of Swahili 19

new. Would it be permissible to consider notions such as, say, exposition and logical
predicate as discussed in Wegener as a kind of forerunner for distinctions such as given
and new or topic and comment? Can the concept logical subject as "a fixed term of
grammar" be equated with a notion like agent?
An important point which must not be overlooked in any such comparison is that one
prominent feature of much of 19th century linguistics is "that for 18th and 19th century
French and German general grammar [...] the existence of thought-counterparts of sen-
tence elements was not an empty article of faith but a living reality, which provided the
notions with some real domain, detached from the empirical-linguistic domain."
(Elffers-van Ketel 1991: 200-201). This means "that grammatical categories were as-
sumed to correspond to mental categories, grammatical systems to systems of thought,
word meanings to concepts and sentences to thoughts." (Elffers 1999:305). Graffi
(2001:15–72) accordingly speaks of "psychologistic syntax" – and, incidentally, men-
tions Wundt as one of its representatives. A notion such as subject thus had quite a dif-
ferent function compared with its modern use: it was applied to describe not only a
grammatical, but at the same time a cognitive concept. This can be seen clearly in We-
gener who uses the term subject not to describe a constituent, or a string of words etc.
but instead a "group of ideas".
This point needs to be taken into consideration in any equation or comparison of
seemingly similar concepts from 19th century and modern linguistic theory.
Coming back to the problem at hand, i.e. the conceptual shift that case and case-
related notions underwent: it has been shown that case-forms such as nominative, accu-
sative etc. were equated with subject, object etc. These notions stood for a variety of
different types of relation at the time. To the degree that post-structuralist linguistics
elaborated and specified these relations and developed a special terminology for them,
case and case-related notions were in some way liberated from conveying this infor-
mation and can therefore be used in a more restricted way to first and foremost refer to
formal, inflectional characteristics.

8. Summary and Conclusion

The putative identity of grammatical concepts used during colonial and modern times
(as suggested by the application of identical labels) ought not to obscure the fact that the
content associated with specific terms may have undergone a conceptual shift.
Such a shift seems to have taken place with respect to the concept case and related
notions such as nominative, accusative etc.: the application of these concepts in gram-
mars of Swahili during colonial times differs substantially from present-day approaches,
where case is no descriptive category of Swahili.
Interestingly, though, a comparison of standard definitions of case from the period
under consideration and modern terminological dictionaries gives no indication that
20 Susanne Hackmack

such a shift has taken place. A closer look at the grammars and the theoretical discus-
sion of the period shows that it is in fact the constructs defining the functional relations
expressed by (morphological) case that have undergone considerable changes.
In other words, the interpretation of grammatical functions such as subject, object etc.
during the 19th and the beginning of the 20th century is different, which itself is a result of
the elaboration and specification of various other types of relation such as semantic roles or
pragmatic functions, which used to be conflated with grammatical functions in the period
under consideration.
Consequently, the interpretation of case-concepts has changed as well: case-forms were
closely associated, in some cases equated with grammatical functions, even to the extent that
the morphological aspect of case was disregarded. Therefore, any modification of the inter-
pretation of grammatical functions would necessarily lead to a modification in the interpre-
tation of case-related notions. Whilst case-notions were used not only to identify inflectional
paradigms of the noun but also to identify various universal types of relation within a sen-
tence, modern linguistics has developed a separate inventory of terms for these universal
types of relation and interprets case-notions as primarily morphological entities. This ac-
counts in part for the seemingly contradictory use of case-notions in Swahili grammars of
colonial times as well as for the difference in the application of case-notions then and now.
In order to arrive at this conclusion, a cursory look at the grammars of colonial times
is insufficient. Instead, the linguistic environment of authors of colonial grammars has
to be taken into account, i.e. their work has to be put into the larger context of linguistic
discussion at the time and reviewed with respect to the influence that individual ap-
proaches and theories may have exerted. Such an analysis reveals that the use of case-
notions in colonial grammars of Swahili is not – at least not in all cases – an indication
of "linguistic imperialism" or "translational grammar" but very much in line with a
wide-spread practice of the period in Germany in which case-notions would denote a
multitude of different relations and be used completely independently of morphological
features.

References
Amidu, Assibi A. (2004): Kiswahili Language Description and Translational Grammar, in: Taiwan
Journal of Linguistics 2.1, 45-68.
Blumenthal, Arthur L. (1975): A Reappraisal of Wilhelm Wundt, in: American Psychologist 30 (11):
1081-1088.
Brockhaus' kleines Konversations-Lexikon (1911). Leipzig: F. A. Brockhaus.
Brauner, Siegmund (1995): Carl Meinhof und Wilhelm Wundt, in: Sprachkulturelle und historische
Forschungen in Afrika: Beiträge zum 11. Afrikanistentag. Köln: Köppe, 59–69.
Brugmann, Karl (1904): Kurze vergleichende Grammatik der indogermanischen Sprachen. Strassburg:
Trübner.
Büttner, C. G. (1891): Hülfsbüchlein für den ersten Unterricht in der Suaheli-Sprache. Leipzig: T. D.
Weigel Nachfolger.
Case in Selected Grammars of Swahili 21

Cienki, A. (1995): 19th and 20th century theories of case: A comparison of localist and cognitive
approaches, in: Historiographia Linguistica 22, 123-162.
Comrie, Bernard (1978): Ergativity, in: Lehmann, Winfred P.: Syntactic Typology: Studies in the
Phenomenology of Language. University of Texas Press, 329-394.
Creissels, Denis (2000): Typology, in: Heine, Bernd & Nurse, Derek (eds.): African Language Struc-
tures. An Introduction. Cambridge: CUP.
Delbrück, Berthold (1901): Grundfragen der Sprachforschung mit Rücksicht auf W. Wundts Sprach-
psychologie erörtert. Strassburg: Trübner.
Delbrück, Berthold (1907): Synkretismus. Ein Beitrag zur germanischen Kasuslehre. Strassburg:
Trübner.
Delius, Siegfried & Roehl, Karl (1939): Wegweiser in die Suaheli-Sprache. Rügenwalde: Albert
Mewes Nachfolger.
Dixon, R. M. W. (1972): The Dyirbal Language of North Queensland. Cambridge: CUP.
Dixon, R. M. W. (2009): Basic Linguistic Theory. Volume 1: Methodology. Oxford: OUP.
Dryer, Matthew S. & Haspelmath, Martin (eds.) (2011): The World Atlas of Language Structures
Online. München: Max Planck Digital Library. Available online at http://wals.info/. Accessed on
2012-11-08.
Elffers, Els (1999): Psychological linguistics, in: Schmitter, P. (ed.): Geschichte der Sprachtheorie 4.
Sprachtheorien der Neuzeit. Tübingen: Narr, 301- 341.
Elffers-van Ketel, Els (1991): The Historiography of Grammatical Concepts. 19th and 20th-century
changes in the subject-predicate conception and the problems of their historical reconstruction.
Amsterdam/Atlanta (GA): Editions Rodopi B.V.
Graffi, Giorgio (2001): 200 Years of Syntax. A Critical Survey. Amsterdam/Philadelphia: John Benjamins.
Haspelmath, Martin (2010): Comparative concepts and descriptive categories in crosslinguistic stud-
ies, in: Language 86.3, 663–687.
Haspelmath, Martin (2010): The interplay between comparative concepts and descriptive categories. A
reply to Newmeyer, in: Language 86.3, 696–699.
Hennig, Mathilde (2009): Zum deutschen Blick auf grammatische Eigenschaften von Kolonialspra-
chen, in: Warnke, Ingo H. (ed.): Deutsche Sprache und Kolonialismus: Aspekte der nationalen
Kommunikation 1884-1919. Berlin/New York: de Gruyter, 120-142.
Herder's Conversations-Lexikon (1857). Freiburg im Breisgau: Herderʼsche Verlagsbuchhandlung.
Holzweissig, Friedrich W. (1877): Wahrheit und Irrthum der localistischen Casustheorie. Leipzig:
Teubner.
Hübschmann, H. (1875): Zur Casuslehre. München: Theodor Ackermann.
Jungraithmayr, Herrmann & Möhlig, Wilhelm J. G.(eds.) (1983): Lexikon der Afrikanistik. Berlin:
Reimer.
Koerner, Konrad (1989): On the Problem of 'Influence' in Linguistic Historiography. In: Koerner,
Konrad: Practicing Linguistic Historiography. Selected Essays. Amsterdam/Philadelphia: John
Benjamins.
Krifka, Manfred (1995). Swahili, in: Jacobs, Joachim et al. (eds): Syntax. Ein internationales Hand-
buch zeitgenössischer Forschung, 2. Halbband [HSK 9,2] Berlin/New York: de Gruyter, Seiten?.
Lestrade, G. P. (1946): Meinhof's contributions to our knowledge of African Languages, in: Doke et
al.: In memory of Carl Meinhof, in: African Studies 5.2, 73–81.
Marty, Anton (1910): Die "logische", "lokalistische" und andere Kasustheorien. Halle: Max Niemey-
er.
Matthews, P. H. (2007): The Concise Oxford Dictionary of Linguistics. Oxford: OUP.
Meinhof, Carl (1906): Grundzüge einer vergleichenden Grammatik der Bantusprachen. Berlin: Rei-
mer.
22 Susanne Hackmack

Meinhof, Carl (1920) Wilhelm Wundt†, in: Zeitschrift für Eingeborenensprachen 10, 241-243.
Meinhof, Carl (1936): Die Entstehung flektierender Sprachen. Berlin: Reimer.
Meinhof, Carl (1940): Die Sprache der Suaheli in Deutsch-Ostafrika. Berlin: Reimer.
Möhlig, Wilhelm J.G. & Heine, Bernd (1999): Swahili Grundkurs. Köln: Köppe.
Newmeyer, Frederick (2010): On comparative concepts and descriptive categories. A reply to Has-
pelmath, in: Language 86.3, 688–695.
Pierer's Universallexikon der Vergangenheit und Gegenwart oder Neuestes encyclopädisches Wörter-
buch der Wissenschaften, Künste und Gewerbe (1857). Altenburg: Verlagsbuchhandlung Pierer.
Planert, Wilhelm (1907): Die syntaktischen Verhältnisse des Suaheli. Berlin: Süsserott.
Pugach, Sara (2012): Africa in Translation. A History of Colonial Linguistics in Germany and Beyond,
1814-1945. Michigan: University of Michigan Press.
Schischkoff, Georgi (ed.) (1957): Philosophisches Wörterbuch. Stuttgart: Alfred Körner.
Seidel, August (1890): Praktische Grammatik der Suaheli-Sprache. Wien/Leipzig: A. Hartleben's
Verlag.
Seidel, August (1900): Suahili Konversationsgrammatik nebst einer Einführung in die Schrift und den
Briefstil der Suahili. Heidelberg: Julius Groos' Verlag.
Seuren, Pieter A. M. (1998): Western Linguistics: An Historical Introduction. Oxford: Blackwell
Publishers.
Velten, Carl (1904): Praktische Suaheli-Grammatik nebst Wörterverzeichnis. Berlin: Wilhem Ba-
ensch.
Wegener, Philipp (1885): Untersuchungen über die Grundfragen des Sprachlebens. Halle: Max Nie-
meyer.
Welmers, Wim (1973): African Language Structures. Berkley/Los Angeles/London: University of
California Press.
Willems, Klaas (1997): Kasus, grammatische Bedeutung und kognitive Linguistik: ein Beitrag zur
allgemeinen Sprachwissenschaft. Tübingen: Günter Narr Verlag.
Wundt, Wilhelm (1904): Völkerpsychologie. Eine Untersuchung der Entwicklungsgesetze von Spra-
che, Mythus und Sitte. Erster Band: Die Sprache (zweiter Teil). Leipzig: Wilhelm Engelmann.