Full download ebook at ebookgate.

com

The Genitive Case in Dutch and German A Study

of Morphosyntactic Change in Codified
Languages 1st Edition Alan Scott

https://ebookgate.com/product/the-genitive-case-
in-dutch-and-german-a-study-of-morphosyntactic-
change-in-codified-languages-1st-edition-alan-
scott/

Download more ebook from https://ebookgate.com

More products digital (pdf, epub, mobi) instant
download maybe you interests ...

Morphosyntactic Change Functional and Formal

Perspectives Oxford Surveys in Syntax Morphology 1st
Edition Fischer

https://ebookgate.com/product/morphosyntactic-change-functional-
and-formal-perspectives-oxford-surveys-in-syntax-morphology-1st-
edition-fischer/

Jews of the Dutch Caribbean Alan Fredric Benjamin

https://ebookgate.com/product/jews-of-the-dutch-caribbean-alan-
fredric-benjamin/

Relative Clauses in Time and Space A Case Study in the

Methods of Diachronic Typology 1st Edition Rachel
Hendery

https://ebookgate.com/product/relative-clauses-in-time-and-space-
a-case-study-in-the-methods-of-diachronic-typology-1st-edition-
rachel-hendery/

The Study of Signed Languages Essays in Honor of

William C Stokoe 1st Edition David F. Armstrong

https://ebookgate.com/product/the-study-of-signed-languages-
essays-in-honor-of-william-c-stokoe-1st-edition-david-f-
armstrong/
The Linguistics of Laughter A corpus assisted Study of
Laughter talk Routledge Studies in Linguistics 1st
Edition Alan Partington

https://ebookgate.com/product/the-linguistics-of-laughter-a-
corpus-assisted-study-of-laughter-talk-routledge-studies-in-
linguistics-1st-edition-alan-partington/

The old high German diphthongization A description of a

phonemic change Irmengard Rauch

https://ebookgate.com/product/the-old-high-german-
diphthongization-a-description-of-a-phonemic-change-irmengard-
rauch/

Heisenberg and the Nazi Atomic Bomb Project A Study in

German Culture 1998 repr 2020 Paul Lawrence Rose

https://ebookgate.com/product/heisenberg-and-the-nazi-atomic-
bomb-project-a-study-in-german-culture-1998-repr-2020-paul-
lawrence-rose/

Rock Mass Stability Around Underground Excavations in a

Mine A Case Study 1st Edition Yan Xing (Author)

https://ebookgate.com/product/rock-mass-stability-around-
underground-excavations-in-a-mine-a-case-study-1st-edition-yan-
xing-author/

From Ad Hoc to Routine A Case Study in Medieval

Bureaucracy Ellen E. Kittell

https://ebookgate.com/product/from-ad-hoc-to-routine-a-case-
study-in-medieval-bureaucracy-ellen-e-kittell/
The Genitive Case in Dutch and German
 Brill’s Studies in
Historical Linguistics

Series Editor
Jóhanna Barðdal (Ghent University)

Consulting Editor
Spike Gildea (University of Oregon)

Editorial Board
Joan Bybee, University of New Mexico – Lyle Campbell, University of
Hawai’i Manoa – Nicholas Evans, The Australian National University
Bjarke Frellesvig, University of Oxford – Mirjam Fried, Czech Academy of
Sciences – Russel Gray, University of Auckland – Tom Güldemann,
Humboldt-Universität zu Berlin – Alice Harris, University of Massachusetts
Brian D. Joseph, The Ohio State University – Ritsuko Kikusawa, National
Museum of Ethnology – Silvia Luraghi, Università di Pavia – Joseph Salmons,
University of Wisconsin – Søren Wichmann, MPI/EVA

volume 2

The titles published in this series are listed at brill.com/bshl

 The Genitive Case
in Dutch and German
A Study of Morphosyntactic Change
in Codified Languages

By

Alan K. Scott

LEIDEN | BOSTON
This publication has been typeset in the multilingual “Brill” typeface. With over 5,100 characters covering
Latin, ipa, Greek, and Cyrillic, this typeface is especially suitable for use in the humanities. For more
information, please see www.brill.com/brill-typeface.

issn 2211-4904
isbn 978 90 04 18144 1 (hardback)
isbn 978 90 04 18328 5 (e-book)

Copyright 2014 by Koninklijke Brill nv, Leiden, The Netherlands.

Koninklijke Brill nv incorporates the imprints Brill, Global Oriental, Hotei Publishing, idc Publishers and
Martinus Nijhoff Publishers.
All rights reserved. No part of this publication may be reproduced, translated, stored in a retrieval system, or
transmitted in any form or by any means, electronic, mechanical, photocopying, recording or otherwise,
without prior written permission from the publisher.
Authorization to photocopy items for internal or personal use is granted by Koninklijke Brill nv provided
that the appropriate fees are paid directly to The Copyright Clearance Center, 222 Rosewood Drive, Suite 910,
Danvers, ma 01923, usa. Fees are subject to change.

This book is printed on acid-free paper.

’t Is waar, de meeste spraakkunstenaars denken anders over de zaak
dan wij, en niet zonder reden; want, door veel over die naamvallen te
praten, maken zij hun boeken zoo veel langer, hun spraakkunst zoo veel
omslachtiger, en hun werk zoo veel mooier, althans zoo veel geleerder.
van lennep 1985 [1865]: 41

…
Als Gymnasiasten wurde uns beigebracht, der Genitiv sei der Kasus der
Gebildeten, das gelte für das Deutsche ebenso wie für das Lateinische.
Noch heute schwingt ein nostalgisches Bedauern mit, wenn vom Tod des
Genitivs die Rede ist.
eisenberg 2008

…
Der (falsch gebrauchte) Genitiv ist der abgespreizte kleine Finger beim
Mokkatässchenhalten.
michael skasa [in sedlaczek 2009]

∵
Contents

Preface ix

1 Introduction 1

2 Morphosyntactic Change 5
2.1 Introduction 5
2.2 Case 5
2.3 Number and Gender 9
2.4 The Effects of Language Change on Case Systems 16
2.5 The Popular Reception of the Manifestations of Morphosyntactic
Change 22
2.6 Codification 26
2.7 Summary of Chapter 2 27

3 The Genitive Case 29

3.1 Introduction 29
3.2 A Typology of the Genitive 29
3.3 The Genitive in the Germanic Languages 36
3.4 The Constructions 44
3.5 Summary of Chapter 3 50

4 Data and Methodology 51

4.1 Introduction 51
4.2 Theoretical Basis 51
4.3 Data Sources 64
4.4 The Databases 91
4.5 Summary of Chapter 4 92

5 The Dutch Genitive 95

5.1 Introduction 95
5.2 The Genitive in Old and Middle Dutch 103
5.3 The Precept: Standardisation and the Genitive Case 110
5.4 The Genitive Case in 16th–19th Century Dutch: A Diachronic
Portrayal 116
5.5 The Genitive Case in Modern Dutch: A Synchronic Portrayal 159
5.6 Other Surviving Remnants of Genitive Morphology 198
5.7 Conclusion 206
5.8 Summary of Chapter 5 208
viii contents

6 The German Genitive 209

6.1 Introduction 209
6.2 The Genitive until the 17th Century 211
6.3 The Genitive Case in 17th–19th Century German: A Diachronic
Portrayal 225
6.4 The Genitive Case in Modern German 249
6.5 Exapted Fragments of the Genitive Case 278
6.6 The Precept: Standardisation, Prescriptivism and the Genitive
Case 295
6.7 Conclusion 307
6.8 Summary of Chapter 6 309

7 Codification and Morphosyntactic Change 311

7.1 Introduction 311
7.2 Morphosyntactic Change Affecting the Genitive in the Data 311
7.3 The Relationship between Codification and Morphosyntactic
Change 316
7.4 A Theoretical Account 321
7.5 Summary of Chapter 7 325

8 Conclusions and Closing Remarks 327

8.1 Introduction 327
8.2 The Findings of the Investigation 327
8.3 Methodological Considerations 337

Appendix 1: Primary Sources 341

Appendix 2: Large Data Tables and Charts 349
References 355
Index 379
Preface

This book reports the findings of my project The Development of the Genitive
in Dutch and German, which was funded by an Early Career Research Fellow-
ship from the Leverhulme Trust, whose support I gratefully acknowledge. The
research was carried out between September 2009 and September 2012 in the
Department of German Studies at the University of Nottingham.
It is probably necessary at the outset to provide some justification for the
publication of yet another book about the genitive case in the Germanic lan-
guages, which is surely one of the most written-about topics in linguistics. I
believe, however, that this book fills a gap in the existing research. The ini-
tial inspiration for my research into the genitive in Dutch and German lies
in my experience of working as Postdoctoral Research Associate on the Arts
and Humanities Research Council-funded project Germanic Possessive -s: An
Empirical, Historical and Theoretical Study (Principal Investigators: Professor
Kersti Börjars and Professor David Denison) at the University of Manchester. In
the course of investigating the possessive -s construction and its prepositional
competitor(s) in English and Swedish, and on becoming acquainted with the
vast body of research into the genitive and possessive -s in English and Swedish
and wishing to make comparisons with other Germanic languages, it became
apparent that there was no comparably detailed, empirically based study of the
development of the genitive case in Dutch and German. This book is an attempt
to fulfil this need.
I am particularly interested in the influence that the standardisation of a lan-
guage can have on the morphosyntactic change affecting that language, and
on the push-and-pull between prescription and actual usage in general: stan-
dardisation was a factor that affected the course taken by the genitive in Dutch
and German, but not in English and the Mainland Scandinavian languages, in
which the genitive had been lost before the advent of standardisation. Further-
more, given the German genitive’s current status as the poster child of perceived
language decline—a perception dating back well over a century but which has
become particularly prominent in recent years—I became interested in mea-
suring up the perceived vulnerability of the genitive against its actual use in
real life sources. A subsidiary aim of this work is, therefore, to make an objec-
tive, empirically founded contribution to the public discourse surrounding lan-
guage “decline”.
This book follows the development of the genitive case—and the nature of
the division of labour between the genitive and its competing, synonymous
constructions—in Dutch and German from the end of the medieval period
x preface

up to the present day. It assesses the relationship between an instance of

morphosyntactic change affecting those two languages, and their standardisa-
tion. The sources studied are therefore drawn from standard language
use (rather than from dialects) and cover a variety of registers and media
to allow for pragmatic variation in using the various semantically equivalent
prescribed and non-prescribed constructions. With the aim of making direct
comparisons between Dutch and German, I have focused primarily on early
modern Dutch and modern German on account of the similarity between
their case systems. To this, I added additional data in the shape of, for Ger-
man, a corpus of early modern newspaper texts, a sample of Martin Luther’s
polemical writings, and a sample of late 18th century personal letters and,
for Dutch, a sample of 16th–19th century informal letters and journals, and
modern formal writing and spontaneous speech. Instead of taking samples
at evenly spaced intervals across a given period, I opted to carry out detailed
synchronic studies of the use of the relevant constructions at certain points
throughout the history of early modern and modern Dutch and German, man-
ually coding the data for structural and pragmatic factors in specially con-
structed databases; within the confines of my project, this approach afforded
a depth and richness of detail that would not have been possible had the
investigation been based on regularly spaced samples from across the entire
period.
The research described here has been presented—sometimes as work in
progress; sometimes in more complete form—at conferences, workshops and
seminars in Antwerp, Austin TX, Bloomington IN, Budapest, Canterbury,
Copenhagen, Groningen, Leiden, Liège, Logroño, Milwaukee WI, Nijmegen,
Nottingham, Oldenburg, Oxford, Reading, Sheffield, Tübingen and Vienna. I am
grateful to the audiences at those presentations for productive discussion and
useful suggestions. I also spent two short but productive spells at the library
of the Institut für Deutsche Sprache in Mannheim; I would like to thank the
helpful staff there and to acknowledge the funding of the University of Not-
tingham Research and Knowledge Transfer Board, which paid for my June 2011
visit.
My research benefitted greatly from the welcoming and supportive working
environment of the German Department at the University of Nottingham, for
which I thank my colleagues. Away from Nottingham, Kersti Börjars and David
Denison are responsible for sparking off my interest in the genitive case as an
object of linguistic research. I am very grateful to Jóhanna Barðdal and Spike
Gildea, the series editors of Brill’s Studies in Historical Linguistics, for accepting
my book into their series and for providing useful feedback on the initial draft
of the book, and I am indebted to Brill’s Stephanie Paalvast, and her predecessor
preface xi

Jasmin Lange, as well as Marjolein Schaake, for guiding me through the various
stages from the submission of the initial proposal to the appearance of the final
monograph.
 chapter 1

Introduction

Two important developments that have shaped the Germanic languages have
been their deflection and their standardisation: the central aim of this investi-
gation is to shed light on the nature of the interaction between morphosyntac-
tic change—such as deflection—and standardisation.1 The effect of deflection
on the genitive case in the Germanic languages has been a frequent topic of
study. Most investigations have focused on the emergence of the invariant,
once-only possessive -s marker from a concordial genitive suffix in English
and Swedish (e.g. English: Allen 2008; Rosenbach & Vezzosi 1999; Rosenbach
et al. 2000; Swedish: Norde 1997, 2001a, b, 2006; Askedal 2002, 2003; Börjars
2003). In both of those languages, these changes were complete prior to stan-
dardisation. However, deflection proceeds differently in standardised and non-
standardised languages, as Vezzosi (2000) shows with reference to Dutch. Up
to the 15th century in Dutch, the decline of the morphological genitive case
in favour of an analytic construction proceeded as expected (Weerman & de
Wit 1999) but, from the 16th century onwards, coinciding with the start of the
standardisation of Dutch, the genitive—and, indeed, case morphology as a
whole—underwent an unexpected resurgence (Vezzosi 2000); no such devel-
opment occurred in English or Swedish. Until the early 20th century in Dutch,
case morphology was preserved alongside the synonymous constructions that
had emerged in the course of deflection; this was manifested primarily in for-
mal written language but was also visible to an extent in informal egodocu-
ments (Chapter 5). As in Dutch, case morphology still remained in German
when standardisation began; here, too, the result was that case morphology
was preserved alongside the non-agreeing and analytic constructions which
had appeared in the course of deflection. Unlike Dutch, this situation persists
today in Standard German; even the dialects, which are not standardised, retain
a case system of sorts, varying in nature from dialect to dialect. Besides Vez-
zosi (2000), little research has been carried out into how deflection affects a
standardised language or, indeed, a language that is still in the early stages of
standardisation. The present research, therefore, is an attempt to fill this gap by
addressing the question of how standardisation affects deflection (and, for that
matter, how deflection informs the standardisation of a language). The tension
between standardisation and deflection is central to this work.

1 The nature of deflection is explored further in Section 2.4.1.

© koninklijke brill nv, leiden, 2014 | doi: 10.1163/9789004183285_002

2 chapter 1

Early modern Dutch and modern German are ideal languages in which to
investigate this tension. The case system of each language already bore the
marks of deflection at the time of the first moves towards standardisation. The
genitive case in both languages had been reduced to an almost exclusively
adnominal case, its adverbal and adjectival roles having all but been lost during
the medieval period. Nonetheless, in contrast to English and Swedish, concor-
dial case morphology was sufficiently intact for it to be included by the early
grammarians—under the influence of Latin—in their prescribed norm. This
support assisted the survival of the genitive case—and case morphology in
general—and set Dutch and German apart from their fellow Germanic lan-
guages (apart from Icelandic and Faroese). Nonetheless, evidence from both
Dutch and German presented in this book indicates that the genitive case had
begun its resurgence before explicit standardisation (i.e. prior to the appear-
ance of codifying works by grammarians, widespread language teaching, and
so on). In Dutch, the loss of case morphology was only postponed; it finally
disappeared—with one exception (Scott 2011a; Chapter 5)—from formal writ-
ten usage during the early part of the 20th century. In German, in contrast,
a case system remains in the standard language and, in reduced form, in the
dialects. Standardisation does not halt or undo deflection, however, and the
genitive case in both languages continued to face competition from alternative
analytic and non-agreeing constructions.
In this investigation the term competition (and the denoting of a con-
struction as a competitor of another construction) does not imply a value
judgement of the type found in the lay linguistic and prescriptive literature,
particularly that produced in German since the 19th century. Rather, compe-
tition denotes the existence—and potential interchangeability—of two or
more semantically equivalent constructions (see Fischer 1987: 274); accord-
ingly, two (near-) synonymous constructions are competitors. By focusing
on the competition between equivalent constructions—the original, synthetic
genitive case on the one hand, and the more recent, analytic constructions that
emerged as a result of deflection on the other—we can gain an insight into
the effects of deflection within a standardised language (in the case of modern
German) or a language in which standardisation is starting to take effect (early
modern Dutch).
The interaction between deflection and standardisation involves various
parameters of variation. This investigation is therefore not simply a study of
the morphosyntactic characteristics of the genitive case in Dutch and German;
rather, it addresses the interfaces of morphosyntax with semantics, sociolin-
guistics and pragmatics. It is constructed from synchronic analyses of the sit-
uation at various points in time from the early modern period to the present
introduction 3

day. These analyses are arranged chronologically in order to give a diachronic

portrayal of the changes taking place.
The genitive case has been chosen as an object of study because of the types
of variation and change that are known to have affected it across the Germanic
languages. Not only has it been joined by a number of semantically equivalent
competing constructions in the course of deflection, but its use has also been
characterised by a great deal of variation. In early modern Dutch and modern
German, the division of labour between the genitive and its competitors varies
according to register and medium. Furthermore, in both languages, the use of
the genitive case has, at different points in time, been the object of prescrip-
tivists’ attention. Thus, the diachronic development and synchronic use of the
genitive and its competitors involve a number of structural and pragmatic vari-
ables, which this investigation addresses. In view of the variables involved, a
large-scale empirical study such as this one is necessary. Accordingly, this work
aims to provide an analysis—with a broad, representative diachronic empiri-
cal basis—of the development of the genitive case in early modern Dutch and
modern German, and to supplement the work on the early periods of English
and Swedish by providing a contrastive view of the same morphosyntactic
changes taking place in standardised languages. Accordingly, while the devel-
opments affecting the genitive in English and Swedish occurred too early to
show the interaction between morphosyntactic change and standardisation,
the present study is able to shed light on this interaction.
This monograph is a tentative attempt to apply a usage-based, construction
grammar framework to language change, and to treat standardisation as a
usage factor and to integrate it into the theoretically founded conclusions
drawn on the nature of morphosyntactic change in standardised languages.
The following research questions are addressed here:

– How does the standardisation of a language affect ongoing deflection in that

language?
– When synonymous constructions are preserved alongside each other
through standardisation, what are the structural and pragmatic factors that
govern their use?
– How similar are the parallel developments that have affected the genitive
case in the history of Dutch and German?
– Can a particular construction be associated with a particular pragmatic
valency (e.g. conventionalisation in a particular register or medium)?

Accordingly, structural and pragmatic analyses are combined here. The relative
balance between structural and pragmatic analysis throughout the investiga-
4 chapter 1

tion depends on the importance of structural and pragmatic factors in the use
of the different constructions; for example, when dealing with the distribution
of the adnominal genitive and von-construction in modern German, besides
one strong structural restriction constraining the genitive, it is mainly prag-
matic factors which govern the use of the two constructions. In contrast, when
dealing with the extension to the German possessive -s construction, more
emphasis is placed on structural factors as these are central (alongside certain
pragmatic factors) to whether or not the construction is used.
The nature of morphosyntactic change—the fundamental process at work
in the developments investigated here—is considered in Chapter 2, in which
the relevant key concepts are defined. This leads into a more detailed typologi-
cal look at the genitive case in Chapter 3 which highlights the shared character-
istics of the Germanic genitives and the genitive case found in other languages
and sets out the main characteristics of the genitive in German and Dutch.
Chapter 4 introduces the theoretical background to the present work; it also
describes the nature of the data collection and the sources used. Chapters 5
and 6 present analyses of the genitive case in Dutch and German, respectively.
These two chapters have a broadly similar structure and focus on the use of
the genitive case and the division of labour between the genitive and the syn-
onymous competing constructions; the language use studied comes both from
older periods and from the present day. In these chapters the focus is on the
individual language in question; however, where relevant, some typological
comparisons are made (for instance, in Chapter 5, comparing the survival of
case morphology in fixed expressions in Dutch as well as in other Germanic
languages). The findings of Chapters 5 and 6 are brought together in Chap-
ter 7, which takes a contrastive approach to Dutch and German and focuses
on the tension between standardisation and morphosyntactic change. Finally,
conclusions are drawn in Chapter 8: the main findings are summarised, the
effectiveness of the theoretical and methodological approach is appraised, and
suggestions are made as to how the findings could be applied in language teach-
ing and how they can contribute to the public discourse on the German genitive
case. Appendix 1 lists the primary sources used in the research presented here
while Appendix 2, in the interest of reader-friendliness, contains some of the
larger tables and figures, which were removed from the main body of the book.
 chapter 2

Morphosyntactic Change

2.1 Introduction

This chapter focuses on the key theoretical concepts relating to the diachronic
changes affecting morphology and syntax and, in particular, how they affect
a language’s morphological case marking. In conjunction with the typological
portrayal of the genitive case given in Chapter 3, the purpose of this chapter
is to provide a foundation to the research described in Chapters 5 and 6 and
to position this research alongside existing work on morphosyntactic change.
First, the concept of case is discussed in Section 2.2; on account of its rele-
vance to concordial case marking, lexical gender is dealt with briefly in Section
2.3. Then, Section 2.4 provides a survey of the effects of language change on
morphological case systems. The matter of the lay linguistic attitudes towards
the synchronic manifestations of morphosyntactic change—of particular rele-
vance when considering the status of the genitive case in modern German—is
introduced in Section 2.5 and finally, in Section 2.6, the tension between the
codification of a language and the morphosyntactic change affecting that lan-
guage is introduced.

2.2 Case

Given the aim of the present study to concentrate on the effects of language
change on morphological case system—specifically, on the competition faced
by case-marking constructions from other constructions (both case and non-
case)—rather than to investigate the theoretical nature of case (or Case), only
a brief summary of the most relevant points of the nature of case is provided
here. The concept of case is defined by Blake (2001: 1) thus:

Case is a system of marking dependent nouns for the type of relationship

they bear to their heads. Traditionally the term refers to inflectional
marking, and, typically, case marks the relationship of a noun to a verb
at the clause level or of a noun to a preposition, postposition or another
noun at the phrase level. [emphasis in original]

Butt (2006: 4) gives a similar definition:

© koninklijke brill nv, leiden, 2014 | doi: 10.1163/9789004183285_003

6 chapter 2

[C]ase is a handy tool for marking semantic relationships between nouns

and verbs, or, more generally between dependents and a head.

In German, for example, nominative marking on a noun phrase marks that

noun phrase as the subject of a verb while accusative marking indicates that
the noun phrase is the object of the verb (1a). Should the constituents of
the sentence become scrambled (within the bounds of what is possible in
German), but the case marking remains the same, the meaning of the sentence
is unaltered (1b). In a language without case marking, such as English, such
scrambling would cause a change in meaning.

(1) a. Der Mann isst den Lachs.

the.masc.sg.nom man eats the.masc.sg.acc salmon
‘The man eats the salmon.’

b. Den Lachs isst der Mann.

the.masc.sg.acc salmon eats the.masc.sg.nom man
‘The man eats the salmon.’

Case marking on a noun phrase can mark that noun phrase as the complement
of an adposition (2). Certain German adpositions permit complements of more
than one case, each case being associated with a particular sense; in ‘in’, for
example, denotes movement with an accusative complement (3a) and stativity
with a dative complement (3b).

(2) Sie lächelte während ihres Auftritts.

she smiled during her.masc.sg.gen performance.masc.sg.gen
‘She smiled during her performance.’

(3) a. Ich laufe in mein Haus.

I run in my.neut.sg.acc house
‘I run into my house.’

b. Ich laufe in meinem Haus.

I run in my.neut.sg.dat house
‘I run in(side) my house.’

Case is indicated by means of affixes and whole words; these, respectively, are
referred to as case markers and case forms by Blake (2001: 2). In a break with
this tradition and in view of the lack of necessity of making the distinction
morphosyntactic change 7

between the two groups in the present study, any item marking case—whether
an affix or an entire word—will be referred to as a case marker. This follows
Moravcsik’s definition of case marker:

A case marker is a formal device associated with a noun phrase that signals
the grammatical role of that noun phrase.
moravcsik 2009: 231, my emphasis

The formal devices in Moravcsik’s definition include bound elements (affixes

and clitics), phonological changes (modification of the stem), and suppletion
(i.e. the use of a particular whole word). Cross-linguistically, morphological
case markers are mostly suffixes (Spencer 2009: 185). Moravcsik’s group of case
markers is broader than that of the present investigation: she also includes
adpositions (Moravcsik 2009: 231 and passim; compare also de Wit’s [1997]
classification of the Dutch preposition van ‘of’ as a genitive marker). While
this investigation does deal with adpositions that denote “case-like” mean-
ings and perform functions otherwise performed by morphological case mark-
ing, adpositions are not treated as case markers here: this would obscure the
move away from synthetic case marking in Dutch and German, which is the
focus of the investigation. Thus, in the course of this book, reference is made,
for example, to genitive markers; that is, to affixes and words that indi-
cate that the noun phrase in which they occur is genitive. The case mark-
ers in German and early modern Dutch are fusional, combining informa-
tion on number, gender and case in a single marker; this is indicated in the
glosses of examples (1) to (4). In addition, both languages display a great
amount of syncretism with regard to the form of their case markers. For exam-
ple, the German determiner den, glossed in (1) as ‘the.masc.sg.acc’, is also
used for ‘the.plu.dat’; mein was glossed in (3a) as ‘my.neut.sg.acc’, but is also
‘my.masc.sg.nom’; the adjective ending -en occurs in all syntactic contexts
except nominative singular (all genders), and feminine and neuter accusative
singular.
Case marking in German and early modern Dutch—in common with, for
example, Latin—is concordial. That is to say, case is marked on certain
words in the noun phrase; mostly on determiners and adjectives, but also on
certain masculine and neuter singular nouns, and the markers vary—and agree
with one another (hence concordial)—according to the gender and number
of the noun. Generally in German and early modern Dutch the head noun
in the genitive noun phrase remains uninflected. In German, the ending -en
attaches to so-called “weak” masculine nouns in the singular in the accusative,
genitive and dative cases (4a), while the “strong” masculine and neuter singular
8 chapter 2

nouns receive an -(e)s ending in the genitive (Section 3.4). This nominal gen-
itive -(e)s also occurred in older Dutch (4b). The concord or agreement
holding within the noun phrase in German and older Dutch is central to this
investigation.

(4) a. Die interessierte Dozentin sprach mit

the.nom.sg.fem interested.nom.sg.fem lecturer.fem spoke with
dem fleißigen Studenten
the.dat.sg.masc hard-working.dat.sg.masc student.dat.sg.masc
‘The interested lecturer spoke with the hard-working student.’

b. de verbeelding des volks

the imagination the.gen.sg.neut people.gen.sg.neut
‘the people’s imagination’ (formal written Dutch, 19th century)

Thus, reference is made throughout this investigation to the concordial gen-

itive, namely the genitive case involving concord within the genitive noun
phrase (i.e. markers varying according to the number and gender of the noun
involved, as in (4b)), as distinct from the possessive -s construction (defined
presently in Section 3.4), which involves the attachment of an invariant marker
(-s) just once at the right-edge of the noun phrase.
The inventory and arrangement of cases in a language constitutes that lan-
guage’s case system:

A case system is defined as the sum total of similarities and dissimilarities

of meanings between the cases, i.e., the sum total of oppositions between
the cases.
de groot 1956: 188

By case system I understand a system by which NPs are marked morpho-

logically to indicate their grammatical functions.
faarlund 1990: 137

A language with a case system is a case language. While it might be uncon-

troversial to describe Latin as a case language, and modern English or mod-
ern Dutch as non-case languages, German is more problematic in this respect.
While it undoubtedly uses inflectional case marking to denote the seman-
tic relationships between the constituents of a sentence, constituent order
(albeit far less rigid than, say, in English) is relatively constrained and syntac-
tic roles are indicated through a combination of morphological marking and
morphosyntactic change 9

constituent order. On this basis, it has been suggested that German should no
longer be regarded as exclusively a case language (Marschall 1998: 37). Further-
more, any account of case in modern German—or, for that matter, written early
modern Dutch—must also allow for the existence of variation, with case some-
times going unmarked on particular members of the noun phrase, such as the
pervasive omission of the nominal genitive -(e)s studied in Chapter 6, or the
occasional omission of the nominal -en ending from weak masculine nouns
(Thieroff 2003; Dovalil 2006: 89–91).

2.3 Number and Gender

In concordial case marking such as that exemplified in (1) to (4), the case
markers vary according to the number and gender of the nouns involved.
The singular-plural number dichotomy found in the Germanic languages is
straightforward and needs little introduction beyond some notes on diachronic
changes affecting it (2.3.1). The nominal gender system of the Germanic lan-
guages is of greater importance to this study as the weakening or loss of con-
cordial case marking is often concurrent with the loss of gender distinctions
and must therefore be taken into account when considering the processes and
effects of morphosyntactic change; this is the topic of 2.3.2.

2.3.1 Number
The Germanic languages all make a singular-plural number distinction. In
English the distinction is marked on nouns (e.g. boot—boots) and on some de-
terminers (e.g. this—these, but the—the); in Dutch the distinction is marked on
nouns (e.g. speler ‘player’—spelers ‘players’) and there is a distinction on neu-
ter determiners (e.g. het ‘the.neut.sg’—de ‘the.plu’), while for common nouns
the singular and plural determiners are identical (e.g. de ‘the.common.sg’,
‘the.plu’; deze ‘this.common.sg’, ‘these.plu’). German, which also marks the
distinction on nouns (e.g. Apfel ‘apple’—Äpfel ‘apples’) has a set of plural deter-
miners (e.g. die ‘the.nom/acc.plu’, diese ‘these.plu’), but they are often syn-
cretic with feminine singular determiners. In all three languages, therefore,
only the entire noun phrase can reveal whether it denotes a singular or plu-
ral referent. In German and older Dutch the plural determiners are identical to
the genitive/dative feminine singular determiners (5),1 only becoming unam-
biguous when combined with a noun (6).

1 The lists in (5) are illustrative, not exhaustive.

10 chapter 2

(5) German:
der ‘the.gen/dat.fem.sg’, ‘the.gen.plu’
dieser ‘these.gen/dat.fem.sg’, ‘these.gen.plu’
meiner ‘my.gen/dat.fem.sg’, ‘my.gen.plu’

Middle Dutch:
der ‘the.gen/dat.fem.sg’, ‘the.gen.plu’
dier ‘this.gen/dat.fem.sg’, ‘this.gen.plu’
sijner ‘his.gen/dat.fem.sg’, ‘his.gen.plu’

(6) German:
die Ansichten der Regierung
the views the.gen.fem.sg government.sg
‘the government’s views’

die Ansichten der Regierungen

the views the.gen.plu government.plu
‘the governments’ views’

This dichotomy has remained generally stable in the face of the numerous
structural changes affecting the Germanic languages. One development of
some (albeit limited) relevance to the present investigation, inasmuch as it
affects the determiner selected to accompany a particular noun, is the treat-
ment as plural of lexically singular nouns denoting a collective human referent
(“team nouns” in the terminology of Lass [1987: 147]; “corporate” nouns in the
terminology of Corbett [2001: 188]). This number mismatch arises through the
singular noun’s semantic characteristics; the usage is particularly entrenched
in British and New Zealand English, but is generally unacceptable in Ameri-
can English (Lass 1987: 147; Corbett 2001: 189). In British English the variation
is pragmatically relevant: the plural variant may be the more natural one, as
in (7a) (in which an attested singular American example is contrasted with
an attested plural British example), or the more informal-sounding one (7b).
When reference is made to sports teams and places, the variation is semanti-
cally relevant, with the singular variant generally denoting the place and the
plural variant generally denoting the team (7c).2

2 The plural variant could also acceptably denote the place viewed as a collection of its
inhabitants; for instance, Do you think Scotland are likely to vote for independence?. The use of
the singular variant to denote a team would sound less natural—but not unacceptable—in
modern British English.
morphosyntactic change 11

(7) a. singular:
The staff is working like mad to process over 600 used rock CDs and get
them out on the floor today …
twitter, usa, 30.5.12

plural:
The staff are having an awesome tip day.
twitter, uk, 27.5.12

b. singular:
Is it just the Queen’s travel plans for the day that the BBC is obsessed with
or everyone’s?
twitter, uk, 1.6.12

plural:
Everyone saying how bad the BBC are doing at this jubilee & how good
Sky are, nothing on ITV? I take it they are showing Bond movies instead.
twitter, uk, 5.6.12

c. singular:
Yes, Scotland is as good as any other nation—in the Union or elsewhere.
[i.e. the country]
twitter, uk, 27.5.12

plural:
Looks like Scotland are playing with the wind at the moment. [i.e. the
team]
twitter, uk, 5.6.12

In German and Dutch, the norm demands that a singular noun with a collective
human be treated as singular. Thus (8b) and (8c), parallel to (8a), are unaccept-
able:

(8) English:
a. The government have decided that …

Dutch:
b. *De regering hebben besloten dat …
the government.sg have decided that
‘The government has/have decided that …’
12 chapter 2

German:
c. *Die Regierung haben entschieden, dass …
the government.sg have decided that
‘The government has/have decided that …’

Instead, the non-mismatched variants in (8b)ʹ and (8c)ʹ are used:

Dutch:
b.ʹ De regering heeft besloten dat …
the government.sg has decided that
‘The government has/have decided that …’

German:
c.ʹ Die Regierung hat entschieden, dass …
the government.sg has decided that
‘The government has/have decided that …’

Nonetheless, the development is attested in those languages, too.3 It can occur,

for instance, when reference is made to pop groups (9) whose name is a singular
noun. Examples such as (9) seem to be restricted to informal language use.
Notice that, as shown in (9a), the use of plural inflection appears to be restricted
to proper names; the collective noun Band later in the example, with the same
referent, takes a verb with singular inflection.

(9) German:
a. Karpatenhund sind zurück— mit neuem Album und neuem
Karpatenhund are back with new.dat album and new.dat
Sound. Die Kölner Band hat sich die musikalischen
sound the Cologne band has refl the musical
Siebenmeilenstiefeln angeschnallt
seven-league-boots strapped-on
‘Karpatenhund are back—with a new album and a new sound. The
band have strapped on their musical seven-league boots’
(Attested at http://www.korg.de/artists/karpatenhund.html [last ac-
cessed 28.2.13])

3 Whether this is as a result of English influence is beyond the scope of the present study.
morphosyntactic change 13

(Belgian) Dutch:
b. Mintzkov hebben het gewoon, er zit toekomst in deze band
Mintzkov have it just there sits future in this band
‘Mintzkov just have it, there is future in this band’
(Attested at http://s4.invisionfree.com/Festivalnoise/ar/t371.htm [last
accessed 28.2.13])

Thus, just like case (2.2) and gender (presently in 2.3.2), number is also affected
by diachronic changes to the morphosyntactic system of a language. Nonethe-
less, these changes are relatively limited and will play only a small part in this
study.

2.3.2 Gender
In a concordial case language such as German or Middle Dutch, knowledge
of the gender of nouns is vital in order to select the appropriate case markers
for use with a particular noun in a particular situation (van der Wal & van
Bree 2008: 133). Originally the Germanic languages had a three way gender
distinction of masculine, feminine and neuter. This distinction remains in
modern German (both standard and dialectal), and in modern Icelandic (10).
Gender distinctions are not made with plural nouns; that is to say, there is a
single plural determiner for use with nouns of all genders.

(10) German:
masculine: der Wagen ‘the.masc.nom.sg car’
feminine: die Freiheit ‘the.fem.nom/acc.sg freedom’
neuter: das Singen ‘the.neut.nom/acc.sg singing’

Icelandic:
masculine: hesturinn ‘horse-the.masc.nom.sg’
feminine: áin ‘river-the.fem.nom.sg’
neuter: borđiđ ‘table-the.neut.nom.sg’
(from Guðmundsson 1922: 83)

In German, this three-way distinction remains sound although some nouns

vary with regard to their gender within the standard norm (11a), across the
German-speaking area (11b), or between everyday language use and specialist
language use (11c) (Durrell 2002: 12; Duden 2005: 153, 228–254).

(11) a. Biotop ‘biotope’: masculine and neuter

Elastik ‘elastic’: neuter and feminine
Zubehör ‘accessories’: neuter and masculine
14 chapter 2

b. Dispens ‘dispensation’: masculine in Germany, feminine in Austria

Spargel ‘asparagus’: masculine in Germany, feminine in Switzerland
Taxi ‘taxi’: neuter in Germany, masculine in Switzerland

c. Filter ‘filter’: masculine in everyday language, neuter in specialist lan-

guage
Meteor ‘meteor’: masculine in everyday language, neuter in specialist
language
Mündel ‘ward’: neuter in everyday language, masculine in legal lan-
guage

Elsewhere in the Germanic languages, the lexical gender distinctions were lost
over time. The most extreme example is English, which no longer makes a gen-
der distinction in its determiners; masculine and feminine natural gender are
distinguished in pronouns with animate referents (i.e. referents that are biolog-
ically masculine or feminine). Feminine pronouns are sometimes encountered
with referents with no biological gender, such as vehicles (particularly boats
and locomotives) and countries. The mainland Scandinavian languages are less
extreme in this respect, retaining a two-gender distinction between common
gender (a conflation of the former masculine and feminine genders) and neuter
gender. As in English, masculine and feminine pronouns continue to refer to
referents with masculine and feminine natural gender, respectively.
The Dutch gender situation is more complex than that of the other Germanic
languages. Given its relevance to the early grammarians’ (temporarily) success-
ful prolongation of the life of the concordial case system (van der Wal & van
Bree 2008: 241–244, 294–296), the current semantics-based gender assignment
in the modern language (Audring 2006; Kraaikamp 2010), and the present-day
use of the adnominal genitive fragment described in Chapter 5, which suggests
that modern Dutch language users have more knowledge of a grammatical
masculine-feminine distinction than usually assumed (Scott 2012), the Dutch
gender system is considered now in more detail.
Concurrently with the weakening of the case system, the Dutch gender sys-
tem lost its distinction between masculine and feminine grammatical gen-
der, at least in the north of the Dutch-speaking area; in the south—including
Flanders and, in the Netherlands, Brabant and Limburg—the original distinc-
tion remains (Hoppenbrouwers 1983: 3). In northern Dutch the weakening and
eventual loss of the masculine-feminine lexical gender distinction is attested
from the 16th century (van der Horst 2008: 803). Today, the former masculine
and feminine genders together comprise the common gender (de-words), dis-
tinguished from neuter nouns (het-words). Nonetheless, for pronoun selection
there remains a masculine/feminine distinction after a fashion: while some
morphosyntactic change 15

nouns are referred to by a masculine pronoun, others are referred to by a fem-

inine pronoun (Audring 2006; De Vos 2009). When a noun denotes a person,
pronoun selection is guided by natural gender: a noun with a male referent
takes a masculine pronoun and a noun with a female referent takes a feminine
pronoun (rare exceptions include the neuter het kind ‘the child’ and het meisje
‘the girl’). The selection of a pronoun to refer to a noun denoting an inanimate
referent, is often informed by the referent’s semantic characteristics, rather
than their lexical gender (Kraaikamp 2010). Although masculine pronouns are
the ‘default’ for reference to nouns that do not denote (individual) humans
(Booij 2002: 37), masculine and feminine pronouns are found referring not only
to common nouns, but also to neuter nouns (and vice versa, see Audring 2006:
88–89). De Vos (2009) describes the tendency to select a feminine pronoun over
a masculine pronoun when referring to a collective noun. This is illustrated by
the examples in (12) in which the noun stichting ‘foundation’ (i.e. a collective
or corporate noun) is referred to by feminine pronouns (in boldface). Of partic-
ular relevance to the present investigation, which—in Chapter 5—deals with
the continued use of a former feminine determiner (i.e. the genitive marker
der), nouns denoting collective human referents, and some nouns denoting
an abstract concept, may be referred to with a feminine pronoun (Audring
2006: 92; Kraaikamp 2010: 11). In addition, many common gender nouns with
inanimate or collective human referents are marked in (larger) dictionaries as
masculine or feminine (e.g. Van Dale 2005–2008); in such dictionaries, nouns
ending in -age, -de, -heid, -ie, -iek, -ij, -ing, -iteit, -nis, -schap, -st, -te and -uur are
marked as feminine. Thus a masculine/feminine distinction of sorts may be
familiar to some speakers.

(12) a. Sinds december 1998 is de SLAS een officiële stichting;

since December 1998 is the SLAS an official foundation
zij ontvangt steun van de gemeente.
she receives support from the local-authority
‘Since December 1998, the SLAS [Stichting Literaire Activiteiten Soest
‘Foundation (for) Literary Activities, Soest’] has been an official foun-
dation; it (lit. she) receives support from the local authority.’
(Attested at http://www.deweekkrant.nl/pages.php?page=2188862
[last accessed 28.2.13])

b. De stichting heeft haar eigen postzegel.

the foundation has her own stamp
‘The foundation has its (lit. her) own postage stamp.’
(Attested at http://www.barsingerhorn.nl/site/?pageId=274 [last ac-
cessed 28.2.13])
16 chapter 2

Another motivating factor is that the morphological structure of the noun to

which the pronoun refers may influence the selection of a feminine pronoun;
that is to say, if the noun’s suffix associates it with feminine gender (i.e. if it
is one of those listed above as being marked as “feminine” in dictionaries),
then some speakers may choose a feminine pronoun (see Donaldson 1983: 62,
162; Audring 2006: 92; Scott 2012). This factor, which is unexpected in modern
Dutch, is returned to in more detail in Chapter 5.
The early grammars of Dutch—as well as grammars and word lists pub-
lished through to the 19th century—also contained information on nominal
gender because, in order to accurately use the prescribed case morphology
(i.e. to select the case marker appropriate to the particular noun), it was nec-
essary for language users to know the gender of the nouns they were using;
from the 17th century, grammarians attempted to increase language users’ gen-
der awareness although there was disagreement among the grammarians as
to the gender of certain nouns (van der Wal & van Bree 2008: 241–244, 294–
296). In the north of the Dutch-speaking area, the determiner de had come to
be used with masculine and feminine nouns and, accordingly, the gender dis-
tinction between those groups was lost; in the south, in contrast, the originally
accusative -n forms of determiners (e.g. den, eenen) had become reanalysed as
masculine determiners, and a masculine/feminine distinction remained (van
der Wal & van Bree 2008: 242–243). It is observed in Chapter 5 that language
users’ use of an incorrect determiner with a particular noun (e.g. a feminine
determiner with a masculine noun, or a masculine/neuter determiner with a
feminine noun) decreased strongly from the 16th to the 19th century, suggesting
that the grammarians may have had some success, both in promoting “correct”
nominal gender use and in promoting the accurate and consistent use of con-
cordial case morphology in general.

2.4 The Effects of Language Change on Case Systems

A central topic of this study is the effect that language change has had on the
morphological case systems of German and Dutch, with focus on the changes
affecting the genitive case. In this section, therefore, the key concepts relating
to this topic are defined and the most important existing research is reviewed
as a background to the analysis pursued from Chapter 5 onwards.

2.4.1 Drift and Deflection

The history of the development of the morphosyntax of the Germanic lan-
guages is centred on the movement from using synthetic constructions—
morphosyntactic change 17

i.e. ones in which grammatical categories are marked morphologically—to

using analytic constructions, in which grammatical categories are marked
periphrastically, i.e. by individual words. This can be exemplified clearly by
comparing an Old High German phrase with its Middle High German and
modern translations, with each version progressively less synthetic than the
previous one (13).

(13) Old High German:

heiligemo geiste
holy.def.dat.masc.sg spirit.dat.masc.sg

Middle High German:

dem heiligen geiste
the.dat.masc.sg holy.dat spirit.dat.masc.sg

Present-day German:
dem heiligen geist
the.dat.masc.sg holy.dat spirit
‘to the holy spirit’

The synthetic language type and the analytic language type may be viewed as
poles on a cline. An attempt to formalise this cline is made in Figure 2.1.4 The
further to the right one moves on this cline, the fewer grammatical distinctions
are marked by morphological endings, with other means such as individual
words and constituent order taking precedence as indicators of grammatical
categories. On this cline, Latin and the Old and Middle periods of the Germanic
languages would be close to the synthetic pole; as far as the modern languages
are concerned, Icelandic and Standard German would remain close to the syn-
thetic pole, the German dialects would be somewhere in the middle, with the
mainland Scandinavian languages and Dutch closer to the analytic pole, and
French and English even closer still to that pole (Faarlund 2001a: 7). A language
need not be uniformly synthetic or analytic, however. Barðdal (2009: 126–130)
notes that the synthetic ditransitive construction (e.g. English I’ll throw you the
ball) has not been replaced by an analytic alternative in which the indirect
object appears as a prepositional phrase (although I’ll throw the ball to you is

4 The cline proposed in Figure 2.1 is intended as an illustration of the nature of the relative
syntheticity (or analyticity) of various Germanic languages (and stages of those languages).
It is not a firmly defined apparatus.
18 chapter 2

figure 2.1 The place of certain Germanic languages on a cline of syntheticity (or analyticity)

also possible in English), and that, in Icelandic, prepositional objects (analytic)

are not used at the expense of dative direct objects (synthetic). The develop-
ments involving the genitive case studied here involve rightward shifts on this
cline of syntheticity (or analycity). Such a shift is interpreted by Tschirch (1969:
185) as resulting from the speakers’ need for increased transparency, which
analytic marking with its individual lexemes (versus the affixes of synthetic
marking) can offer.
The term drift denotes rightward movement along the cline illustrated in
Figure 2.1 (Vezzosi 2000, passim; Faarlund 2001a: 5):

Drift is understood as a unidirectional series of changes whereby lan-

guages go from being more synthetic to more analytic. The changes that
constitute this drift include loss of morphological case, its function being
taken over by word order, and the increased use of function words.
faarlund 2001a: 5

As languages move rightward, they may undergo deflection:5

It should be noted that the term deflexion, as here defined, refers to the
loss of inflectional categories, not necessarily to the loss of all inflections.
In the case of nominal morphology, for example, it refers to the loss of

5 The spelling deflection is used here instead of the deflexion found elsewhere (e.g. Norde
2001a,b, 2006; Allen 2008) simply because of the former’s orthographic resemblance to in-
flection, and to provide a transparent link to the verb and past participle/adjective deflect
and deflected, respectively, which are also used in this study.
morphosyntactic change 19

case as an inflectional category. This resulted of course in the loss of a

number of individual suffixes, but some of the old endings were retained
as gender/number affixes [cross-reference omitted] or even exapted to
new functions [cross-reference omitted].
norde 2001a: 240; see also norde 2001b: 242

Norde (2001a: 240) characterises deflection as “one of the most fundamen-

tal changes in the history of the Germanic languages”. Indeed, all Germanic
languages have been affected by it to some extent. Modern Dutch, English
and the Mainland Scandinavian languages, with their lack of concordial case
marking and their reduced lexical gender systems, and their reliance on con-
stituent order to mark grammatical relations, are the most deflected—and
most analytic—Germanic languages. Modern Icelandic, in contrast, with its
concordial case marking and rich inventory of inflectional affixes, and its three
genders, is the least deflected (and the most synthetic); nonetheless, its con-
stituent order has become more rigid (a development associated with deflec-
tion) over time (Barðdal 2009: 130). Modern Standard German, as suggested
by its position in Figure 2.1, is more deflected than Icelandic but less deflected
than its fellow mainland Germanic languages: while it retains three genders
and a concordial four-case system, those cases often face competition from
each other or from equivalent non-case constructions, and its inventory of case-
marking affixes (/ə/, /əm/, /ən/, /ər/, /əs/) is both smaller than that of modern
Icelandic and that of the oldest period of German as a distinct language.6
Consequences of deflection within the Germanic languages include the
existence of competition between case constructions and non-case construc-
tions. An ongoing example of this is the competition in German between the
adnominal genitive and an analytic equivalent in which the preposition von
‘of’ marks the “genitive” relationship between the two noun phrases. Com-
pleted examples, in which the analytic construction has prevailed over the syn-
thetic original, include the competition in Dutch between the dative case and
a periphrastic construction with the preposition aen [aan in modern orthogra-
phy] ‘on’ which took place from the 16th century onwards (van der Horst 2008:
801), and the competition in the earliest stages of German between the instru-
mental case and a periphrastic construction with the preposition mit ‘with’
(Keller & Mulagk 1995: 204).

6 The weakening of full vowels to schwa in unstressed syllables, which contributed to the reduc-
tion of the inventory of distinct inflectional affixes, is a key development in the transition from
Old to Middle High German and had consequences for the development of concordial case
morphology in German. It is returned to again in Chapter 6.
20 chapter 2

A move from concordial to once-only marking can occur in the process of

deflection; examples of this are the pervasive omission of the nominal genitive
-(e)s suffix in German (giving, for example, des Wissen ‘the.gen knowledge.∅’
instead of the expected des Wissens ‘the.gen knowledge.gen’), and the devel-
opment of the possessive -s construction, in which possession is only marked
on the right-edge of the possessor noun phrase. Once-only marking of case—
i.e. single encoding—is more efficient, and therefore advantageous to both
speaker and hearer/reader, than the concordial marking that demands agree-
ment markers on every element in the noun phrase (Norde 2001b: 258). This
fits with the principle of minimize forms as formulated by Hawkins (2004:
38):

Minimize Forms (MiF)

The human processor prefers to minimize the formal complexity of each
linguistic form F (its phoneme, morpheme, word, or phrasal units) and
the number of forms with unique conventionalized property assign-
ments, thereby assigning more properties to fewer forms. These mini-
mizations apply in proportion to the ease with which a given property
P can be assigned to a given F.

Within MiF the number of units making up a form can be reduced; for instance,
a long word can be abbreviated, or a pronoun can replace a complex noun
phrase (Hawkins 2004: 38).
As Norde (2001a: 240; see also Barðdal 2009 and Barðdal & Kulikov 2009)
notes, the origin of deflection has been variously attributed to phonological
changes (e.g. the reduction to schwa of full vowels in unstressed syllables), the
weakening of case government by verbs and prepositions, and language con-
tact. Because the origins of deflection predate the periods focused on in the
present study—whose focus is, instead, on the consequences of deflection—
they are not dealt with further here.

2.4.2 The Survival of Case Morphology

Nevertheless, despite the actions of deflection within a language, case mor-
phology may remain. Often, the survivors are former case markers in a now
caseless language, which have been exapted to perform a new function. The
most celebrated example of such a survivor is the possessive -s construction
(a former concordial genitive marker) in English, Mainland Scandinavian and
Dutch; another widespread example in Dutch and Mainland Scandinavian
is the linking element -s- in compounds, which also emerged from the con-
cordial genitive -s. The absence of concordial case is not a prerequisite for
morphosyntactic change 21

such exaptation, however: even standard German, which retains the genitive
-s ending as part of its case system, has a possessive -s and linking -s. A much
rarer occurrence is the survival of a whole chunk of concordial case morphol-
ogy as a productively used construction complete with a determiner-noun
agreement relationship that reflects very closely the agreement relationship
that would have held when the language still had its concordial case system:
such a survivor is the Dutch adnominal genitive construction (Scott 2011a,
2012).
Relics of the lost case system can still be found in English, Dutch and Main-
land Scandinavian in the form of some of their pronouns; this cross-linguis-
tically attested phenomenon does not imply that those languages still have
case (Spencer 2009: 195). For example, in English and Dutch we find pronoun
distinctions of the kind exemplified in (14); English also continues to make a
distinction between a subject and object interrogative pronoun (15), but this is
now an artificially learned distinction made only in prestigious language use
(Lass 1987: 152; Lasnik & Sobin 2000: 344–347).

(14) English:
he ‘3sg.masc.subj’—him ‘3sg.masc.obj’—his ‘3sg.masc.poss’
we ‘2plu.subj’—us ‘2plu.obj’—our ‘2plu.poss’

Dutch:
zij ‘3sg.fem.subj’—haar ‘3sg.fem.obj/poss’
jij ‘2sg.subj’—jou ‘2sg.obj’—jouw ‘2sg.poss’

(15) subject: Who is that?

object: The student whom I used to teach is standing over there.

Entire concordial case systems can also survive in the face of deflection, but
generally need the support of standardisation to do so. Deflection had removed
concordial case morphology from English and Mainland Scandinavian by the
early modern period, but had been slower to take effect in Dutch and German.
The relatively rapid loss of case morphology from English and Swedish is
ascribed by Barðdal (2009: 125, 155) to those languages’ relatively early (i.e.
pre-standardisation) exposure to language contact while German and, most
extremely, Icelandic were less subjected to contact (Dutch is not mentioned).
The contact between northern English speakers of Old English and Danish
speakers of Old Norse led to deflection occurring in northern English dialects
around two centuries in advance of dialects in the south (Jespersen 1912: 81; see
also Kroch 2001). Accordingly, German and Dutch retained a concordial case
22 chapter 2

system, albeit one that was in some disarray (particularly in Dutch). This case
system was therefore familiar to the earliest grammarians, who included it in
their prescribed norm for the standard language. Even before the grammarians’
work had taken effect (or had even taken place), concordial case morphology
was part of the written convention in both languages: evidence of this comes
in the shape of Martin Luther’s 16th century Reformation texts (predating the
earliest German grammars by over a century) and the 17th century Dutch texts
studied here (which are roughly contemporary with the appearance of the
first Dutch grammars). Once their position in the norm was secure, the Ger-
man and Dutch concordial case systems became entrenched and remained in
use, at least in the written languages. The Dutch concordial case system sur-
vived until a spelling reform in the 1940s, while the German system remains
in use to this day, making it an anomaly among the mainland Germanic lan-
guages.

2.5 The Popular Reception of the Manifestations of Morphosyntactic

Change

It is beyond the scope of the present investigation to attempt a comprehen-

sive overview of popular approaches and attitudes to the manifestations of
diachronic change; for German, the reader is referred to Davies & Langer
(2006). Nonetheless, given the often prominent position of concordial case
marking in prescriptive works from the 16th century onwards, an overview
of linguistic prescriptivism and the field of lay linguistics is sketched out
here—with particular reference to morphosyntactic change—as a back-
grounds to the statements made on the relationship between prescription and
developments affecting the genitive (for example, in Sections 5.3 and 6.6, and
Chapter 7).
Linguistic prescriptivism, at its most basic, can simply denote an instruc-
tion by a particular language authority (whether a teacher, grammar book,
journalist, etc.) to use, say, a particular construction in a particular way, or
one particular construction instead of another, synonymous one. More usually,
however, it refers to “[t]he imposition of arbitrary norms upon a language, often
in defiance of normal usage” (Trask 1999: 246). As Davies & Langer (2006: 46)
observe, the prescriptive preference for a particular variant or the stigmatisa-
tion of another has nothing to do with how widely used a variant is; using the
examples of the stigmatised English split infinitive and the use in German of a
dative complement with the preposition wegen ‘because of’, they also note that
prescriptive rules are
morphosyntactic change 23

not necessary for communication to take place on a referential level, but

if these rules are not followed, communication may be negatively affected
in the sense that certain sections of the population will not take the
speaker seriously. From a social-indexical point of view, they are therefore
not necessarily redundant.
davies & langer 2006: 47

A number of the constructions investigated in this study have, at some point

in their history, been either promoted or stigmatised (and, in some instances,
both) by prescriptivists. This investigation therefore straddles an interesting
interface between morphosyntactic change and pragmatics and sociolinguis-
tics.
Ever since the earliest grammarians, everyday language usage has often been
regarded as having a damaging influence on a language. In the 17th century,
the German grammarian Schottel—along with his fellow grammarians of the
time—saw the language use of the uneducated speakers as having harmed a
once ideal language: grammars were seen as a means of rectifying this harm
(Lange 2005: 64). Thus, despite the deflection that would have been clear in
the Dutch and German case systems of the time, the early grammarians of
both languages (e.g. Twe-spraack 1584, Schottel 1663) included a morphological
case system—emulating the Latin paragon as well as keeping alive the system
present throughout the history of the Germanic languages up to that point—as
part of their precept.
While linguistic prescription and the production of prescriptive grammars
are centuries-old, a more recent phenomenon, pervasive in the German- and
Dutch-speaking countries, is lay linguistics:

‘Laien-Linguistik’ ist eine an die breite Öffentlichkeit gerichtete praxisori-

entierte Sprach- und Kommunikationslehre zur Lösung muttersprach-
licher Probleme. Sie ist eine für und bisweilen auch von (gebildeten)
Laien betriebene handlungsorientierte Thematisierung des Gebrauchs
von Sprache in Kommunikation in Form von bestimmten Publikationen
und Lehrangeboten (‘Seminare’, ‘Trainings’).

“Lay linguistics” is a practice-oriented teaching of language and commu-

nication aimed at a broad public for solving problems relating to one’s
native language. It is an action-oriented taking as a theme, run for and
now and then also by (educated) lay people, of the use of language in
communication in the form of specific publications and teaching-events
(“seminars”, “training sessions”).
antos 1996: 13, my translation
24 chapter 2

The most prominent German prescriptivist of recent times is the journalist

Bastian Sick, whose scope covers lexical and pragmatic issues as well as man-
ifestations of morphosyntactic change. When dealing with lay linguistic pre-
scription relating to the genitive, Sick’s work (2004 and later)—which started
off as a column in the news journal Der Spiegel—is particularly relevant, both
in terms of its content (manifestations of diachronic change on morphology
and syntax are the topic of many of his columns and chapters), and, famously,
in terms of its title. The title makes specific reference to the perceived suppres-
sion of the genitive by the dative, and makes jocular use of the periphrastic
possessive construction which is one of the genitive case’s competitors in mod-
ern German: Der Dativ ist dem Genitiv sein Tod ‘the dative is the.dat genitive its
death, i.e. the dative is the death of the genitive’. Although the idea that the
genitive case is in mortal danger (particularly as a result of encroachment from
the dative) is not new (Section 6.6), Sick has certainly contributed greatly to
the perception of a genitive under threat, and his title has entered the popu-
lar consciousness. Nonetheless, the genitive and its competitors are just one of
many manifestations of morphosyntactic change dealt with by Sick (see also
Scott 2011b: 56); among the topics he deals with are the increasing productiv-
ity of the derivational suffix -bar, inflection of loan words (e.g. plurals of loan
nouns, conjugation of English verbs in German), and weak and strong verb
inflection.
The success—or notoriety—of these modern lay linguistic works and others
like them has generated a backlash from linguists and non-linguists alike.
While the value of Sick’s work (which encompasses his book series, newspaper
columns, lecture tours, a DVD, and more) in drawing interest to the German
language and to certain aspects of its use is recognised, criticism of his work
focuses on both its linguistic value (and accuracy) and the tone in which
Sick addresses his audience (or, more precisely, with which he describes the
producers of “inaccurate” language use):

Der Ton, wenn er auch meist geschickt gewählt ist, kommt zuweilen etwas
ironisch und provozierend, bisweilen sogar besserwisserisch daher. Und
das kann sich der Autor nicht immer leisten. Denn neben den guten
und treffenden Darstellungen strotzt das Sick’sche Werk eben auch von
Ungereimtheiten, Unstimmigkeiten, Halbwahrheiten, Pedanterien bis
hin zu reinen und groben Fehlern.

The tone, even if it is mostly adroitly chosen, comes over at times some-
what ironic and provocative, and now and again even in a know-it-all
manner. And the author cannot always afford that. For beside the good
morphosyntactic change 25

and apposite portrayals, Sick’s work is just also full of inconsistencies,

discrepancies, half-truths, pedantries, all the way up to pure and gross
mistakes.
meinunger 2006: 10, my translation

Sicks Publikum scheint weniger aus jenen Leuten zu bestehen, die vor
ihren Reisen die ‘Visas’ beantragen und ‘leckere Pizza’s’ auf die Tafeln
ihrer Imbißbuden schreiben. Vielmehr scheinen es die zu sein, denen
der Unterschied zwischen dem Dativ und dem Genitiv bekannt ist, nor-
mal gebildete Mittelschichtsbewohner, denen Sicks dumme Späße die
angenehme Gewißheit verschaffen, daß es zu denen da unten noch ein
ganzes Stück Wegs weit ist.

Sick’s audience seems to consist less of those people who apply for “Visas”
before their travels and who write “tasty pizza’s” on the boards of their
cafés. Rather, it seems to be those who are already aware of the difference
between the dative and the genitive, normal educated members of the
middle class, for whom Sick’s silly jokes provide the comfortable certainty
of there being a fairly long way to those down below.
seidl 2006, my translation

[The attachment of a plural -s to the plural form Visa ‘visas’ (< Visum
‘visa’), and the use of the apostrophe before plural -s are topics dealt with
by Sick. Whether Seidl’s use of the archaic nominal partitive genitive in
ein ganzes Stück Wegs ‘a whole piece way.gen, i.e. a fairly long way’ is a
reference to Sick’s promotion of the genitive, is unclear.]

Sick’s success is viewed by Klein (2009: 154) as a backwards step towards older
prescriptive attitudes to language use.
The inclusion of genitive-related phenomena in some of the modern Ger-
man prescriptive works means that their pronouncements on the genitive
will necessarily form a part of this investigation. It is therefore necessary—
particularly in view of the existence of the linguistic works that aim to debunk
the lay linguists—to adopt a position on lay linguistic prescriptions regarding
genitive use (and language use in general): for the purposes of this investi-
gation, these lay linguistic works are regarded simply as a source of data on
attitudes towards genitive use.
Lay linguistic prescriptions relating to the genitive case in Dutch and Ger-
man are addressed in detail, and compared to the situation in usage data, in
Chapters 5 and 6, respectively.
26 chapter 2

2.6 Codification

While the efforts of the present-day lay linguistics can be viewed in the context
of trying to protect one’s language from perceived misuse, and of taking partic-
ular usages as markers of certain groups of people (Section 2.5), the work of the
early grammarians, who also shared these aims, must be viewed in a broader
context, namely that of codifying and standardising the languages. While Lange
(2005: 83), for instance, concludes that the 17th century German grammarians
were not the creators of the standard German norms—indeed, the levelling
of dialectal differences by 16th century printers in order to make their prod-
ucts more widely sellable, and Martin Luther’s use of his relatively neutral east
central variety in his Reformation writings and Bible translations, among other
factors, predate the grammarians by over a century (Stedje 2007: 146–154)—and
there is, at most, sparing evidence of the influence of the earliest Dutch gram-
mars in contemporary texts, those grammarians—and their successors—were
involved in the standardisation of their language and in the promotion of cer-
tain constructions and stigmatisation of others.
The example of English and Mainland Scandinavian makes clear that, in the
absence of prescriptive support, a morphological case system can succumb
to deflection and vanish from a language; the same is shown by the loss of
the concordial instrumental case in favour of an analytic alternative in Old
High German. Equally, early modern Dutch and modern German indicate that,
with the support of a standard norm, a concordial case system may remain.
Regardless of their role as creators or assistants, the grammarians and their
work is therefore relevant to the present investigation. The tension between
the standardisation of a language and the deflection going on in that language
is the fundamental phenomenon in the research described here.
Four stages of standardisation are identified by Haugen (1966: 18–24):

1. Selection of norm, i.e. a particular variety is chosen to become the stan-

dard
2. Codification of form, i.e. the standardisation of the structural features
of a language; this involves the production of normative works such as
grammars and dictionaries (Swann et al. 2004: 41)
3. Elaboration of function, including the use of the language beyond every-
day situations, such as scientific and poetic language
4. Acceptance by the community

The stage most relevant to the present study is codification, because this is
the one in which particular variants become promoted and propagated, and
morphosyntactic change 27

others stigmatised. For example, it was at the codification stage of German

and Dutch that a morphological case system was taken as part of the norm
despite the deflectional developments that had affected the system up to that
point. Morphological case marking vanished from written Dutch usage once
it was removed from the code in the 1940s; in contrast, morphological case
marking continues to be promoted as part of the standard German norm, while
various deflection-related constructions are stigmatised. This relation between
codification and deflection is dealt with implicitly throughout Chapters 5 and
6, and is addressed in detail in Chapter 7.

2.7 Summary of Chapter 2

The purpose of this chapter was to explore the nature of the process which
is central to the present investigation, namely morphosyntactic change. The
nature of case and, in particular, concordial case marking were explored, with
number and gender also considered in view of their relevance to concordial
case marking (that is to say, because case markers vary according to number
and gender). Then, the central processes of this investigation were introduced,
namely the reductive effects of language change on case systems (the concept
of deflection, which is in the spotlight throughout this book, was defined), the
survival of case morphology (whether isolated relics or a whole case system),
and the reception of the manifestations of morphosyntactic change by gram-
marians and lay linguists. Finally, the matter of standardisation—particularly
the codification stage—was considered in light of its revelvance to morphosyn-
tactic change: the interface between codification and morphosyntactic change
is the focus of this book. This chapter, along with Chapter 3, which focuses on
the genitive case, provides a basis for the investigation described in Chapters 5
and 6.
 chapter 3

The Genitive Case

3.1 Introduction

This section builds upon the overview of morphological case marking and lan-
guage change given in the previous chapter, and focuses on the genitive case.
The cross-linguistic characteristics of the genitive are sketched in Section 3.2,
showing that the roles performed by the genitive case in the Germanic lan-
guages are much the same as those found throughout the languages of the
world. In Section 3.3 the use and development of the genitive across the Ger-
manic languages is addressed. It is noted that, as far as empirically based
diachronic investigations are concerned, attention has been focused on English
and Swedish. The present investigation is an attempt at providing an equiva-
lent portrayal of the developments in Dutch and German. Finally, in Section 3.4
the constructions referred to throughout this investigation—namely the geni-
tive and the non-genitive “competitors”—are identified, defined and exempli-
fied.

3.2 A Typology of the Genitive

The term genitive comes from Latin: cāsus genetīvus or genitīvus ‘case denot-
ing origin, belonging’ (Bußmann 2002: 246; Duden 2011a: 700). The sense of
the Latin name is reflected throughout the genitive’s various uses to vary-
ing extents. A factor characterising the genitive case cross-linguistically is its
semantic generality; Wellander (1956: 159) observes a “Farblosigkeit” [‘colour-
lessness’] of the (Germanic) genitive. In essence, the genitive denotes a “thing-
to-thing relation” (De Groot 1956: 189). This generality can lead to problems in
classification (Palmer 1961: 290), or at least to divergences between different
scholars’ classifications of the roles performed by the genitive. Latin grammars
typically list upward of thirty types of genitive, but differ from each other in the
exact arrangement and classification (De Groot 1956: 192, 194; Bondzio 1967: 3),
while the number of genitive types in German can vary greatly depending on
the scholar, with Helbig (1973: 210–213) listing 25 adnominal and eight adver-
bal genitive types, Bondzio (1967) listing seven adnominal genitive types and
Duden (2005: 833–839) nine adnominal genitive types; Ballweg (1998), in con-
trast, attempts to provide a single, unified—if abstract—interpretation of the

© koninklijke brill nv, leiden, 2014 | doi: 10.1163/9789004183285_004

30 chapter 3

adnominal genitive. The classifications of the German genitive are returned to

in Section 3.4, where the simplified classification (albeit more complex than
that of Ballweg [1998]) adopted for the present investigation is set out. Seman-
tic emptiness is often noted in definitions and explanations of the genitive
although, often, the possessive sense—typologically, the most widespread role
of the genitive (Bußmann 2002: 246)—is mentioned:

This case is used to signal the fact that one noun is subordinate to the
other, i.e. one noun is the head and the other noun is the modifier which
adds some further specification to the head.
katamba 1993: 240

Case whose basic role is to mark nouns or noun phrases which are depen-
dents of another noun.
matthews 1997: 144

Morphologischer Kasus, der primär zur Kennzeichnung des → Attributs

eines Substantivs dient

Morphological case which primarily serves to indicate the attribute of a

noun
bußmann 2002: 246, my translation

A distinctive case form typically marking a noun phrase which serves a

possessive role within a larger noun phrase. […] The genitive is unusual
among case forms in that it does not normally express an argument or
adjunct of a verb; nevertheless, in languages with well-developed case
systems, it is usually integrated morphologically into the case system.
trask 1993: 118

Possessive relations are usually induced from the context or the lexical
semantics of the corresponding nominals [references omitted]. Thus, (1)
[Russian kniga Lizy ‘book.nom.sg Liza.gen.sg’] could mean “the book
that Liza owns” or “the book that “Liza wrote” or “the book that Liza
photographed” etc., depending on the context […].
lander 2009: 581

A genitive case occurs in many different language families, such as Altaic, Cau-
casian, Dravidian, Indo-European, Semitic and Uralic (Blake 2001: 151). Cross-
linguistically, the genitive is generally an adnominal case—denoting a rela-
the genitive case 31

tionship between two noun phrases—with the genitive relationship marked

on the noun phrase that is the dependent of the head noun phrase within a
larger noun phrase; it is also encountered adverbally, that is, as the case of
the complement of certain verbs (Blake 2001: 98, 151; Lander 2009: 581). Butt
(2006: 8) classes the genitive as a nominal case on account of its being marked
on “nominals which are licensed by other nominals”, as opposed to a verbal
case, which is licensed by verbal predicates. The relative use of the adnom-
inal and adverbal genitive varies between languages and over time within a
single language. The Latin genitive—like the modern German genitive—is pre-
dominantly used adnominally to denote a possessive relationship—in a broad
sense—between two noun phrases and is accordingly often referred to as the
“possessive case” with the genitive-marked noun phrase denoting the “posses-
sor” of the head noun phrase (e.g. Blake 2001: 98, 151; see also Haspelmath 2009:
513) (1). A role of the genitive in Latin and German (but not exclusive to those
languages; cf. also the present-day English possessive -s construction [some-
times also termed “genitive”, e.g. Butt 2006: 8]) is to render adnominally an
adverbal complement (2); the subjective genitive (2b) corresponds to the
subject of the verb (2a), while the objective genitive (2d) corresponds to what
would be the object of the verb (2c) (Blake 2001: 98). In modern German, the
subjective and objective genitives are the two most frequent uses of the adnom-
inal genitive, with the former being a great deal more frequent than the latter
(Chapter 6, Figure 6.18). Canonically, genitive marking appears on the noun
phrase denoting the possessor (Lander 2009: 582); this is shown in the exam-
ples in (1) and (2).

(1) Latin
a. consulis equus
the-consul.gen horse
‘the consul’s horse’ (Blake 2001: 98)

German
b. das Pferd des Konsuls
the horse the.gen consul.gen
‘the consul’s horse’1

1 The formal characteristics of German concordial case marking are dealt with in Section
3.4. Until then, the unspecific gloss ‘gen’ is sufficient to indicate the presence of genitive
inflection.
32 chapter 3

(2) a. Der Mann singt das Lied.

the man sings the song
‘The man sings/is singing the song.’

b. der Sänger des Lieds

the singer the.gen song.gen
‘the singer of the song’

c. Jemand bemisst die Entschädigung.

somebody calculates the compensation
‘Somebody calculates the compensation.’

d. die Bemessung der Entschädigung

the calculation the.gen compensation
‘the calculation of the compensation’

The Latin adverbal genitive was restricted to a small group of verbs, while in Old
English it was much more widespread (Blake 2001: 151). In present-day standard
German only a handful of verbs demand a genitive complement, sixteen being
encountered in the course of the present investigation (Section 6.4.11); which
is far fewer than the circa 260 verbs that assigned a genitive complement in
Middle High German but which, over time, came to take accusative or preposi-
tional complements instead (Ágel 2000: 1870; Paul 2007: 340–341; Sections 6.2.1,
6.2.3 and 6.3.6).
Besides its adnominal and adverbal uses, the genitive also appears cross-
linguistically as the case of complements of certain adpositions, particularly
when these adpositions have emerged from nouns (Lander 2009: 589), for
example:

(3) German:
anstatt einer Möwe
instead a.gen seagull
‘instead of a seagull’ (Dortmund)

Aghul:
Xul-ar-i-n üdih
house.plu.obl.gen front
‘in front of the houses’ (Lander 2009: 589)
the genitive case 33

Language-specific peculiarities notwithstanding, the genitive case is used in

a broadly similar way cross-linguistically. This is exemplified now with three
non-Germanic languages: two Indo-European and one non-Indo-European.
The Latin genitive is principally used adnominally (4a) with a variety of
senses—including a partitive sense (4b)—and is the case assigned by a number
of verbs (4c) and adjectives (4d) (Woodcock 1985 [1959]: 50–60). The generality
of the Latin genitive—common also to its Dutch and German counterparts, as
Chapters 5 and 6 make clear—is reflected in Palmer’s (1961: 294) statement that
“a noun in the genitive defines and delimits the range of reference of another
noun or a verb”.

(4) a. patris amicus

father.gen friend
‘father’s friend’ (Palmer 1961: 290)

b. cadus vini
cask wine.gen
‘a cask of wine’ (Woodcock 1985 [1959]: 53)

c. Miltiades proditionis accusatus est.

Miltiades treachery.gen accused is
‘Miltiades was accused of treachery.’ (Cicero, in Woodcock 1985 [1959]:
56)

d. pauper aquae
poor water.gen
‘poor in water’ (Woodcock 1985 [1959]: 55)

Unlike the Germanic languages, the main alternatives to the Latin adnom-
inal genitive are not a prepositional paraphrase or another case, although
both are possible to an extent, but rather a phrase involving an adjective (4e)
(Woodcock 1985 [1959]: 50). Competition between the adnominal genitive and
an adjectival phrase also occurs in German (Wellander 1956: 156), although
adjectival alternatives such as that in (4f) now often have an archaic charac-
ter.2

2 Furthermore, the adjectival alternative is limited in its type frequency: das väterliche Buch
‘father’s book’ or das väterliche Auto ‘father’s car’ are, at best, dubious.
34 chapter 3

e. fratris mors
brother.gen death
‘a brother’s death’

fraterna mors
brotherly.gen death
‘a brother’s death’ (Woodcock 1985 [1959]: 50)

German:
f. das Haus des Vaters
the house the.gen father.gen
‘father’s house’

das väterliche Haus

the fatherly.neut.acc.sg house
‘father’s house’ (Wellander 1956: 156)

The Polish genitive is used, among other roles, adnominally to denote posses-
sion (5a), adverbally, particularly if the verb is negated (5b), for the comple-
ments of certain prepositions (5c), and partitively (5d) (Bielec 1998: 106). Note
the once-only marking on the forename rather than the surname (in contrast
to the Germanic possessive -s) in (5a).

(5) a. Nowa piosenka Edyty Bartosiewicz na antenie Programu

new song Edyta.gen Bartosiewicz on antenna programme
Trzeciego.
third
‘Edyta Bartosiewicz’s new song (broadcast) on Channel 3.’
(Attested at http://www.edytabartosiewicz.pl/PRASA/WPR800.HTM
[last accessed 28.2.13])

b. Nie lubįe jabłek.

not I-like apples.gen
‘I don’t like apples.’ (Bielec 1998: 26)

c. do Anglii
to England.gen
‘to England’ (Bielec 1998: 97)
the genitive case 35

d. Kupįe mleka i sera.

I’ll-buy milk.gen and cheese.gen
‘I’ll buy some milk and cheese.’ (Bielec 1998: 121)

As in Polish and German, the Finnish genitive denotes possession adnominally

(6a), and is the case of the complements of certain prepositions (6b); it also
occurs adverbally as the subject of certain verbs (6c).

(6) a. Kirjojen sisältö on muuttunut.

the-books.gen content has changed
‘The content of the books has changed.’ (Karlsson 1999: 95)

b. talon sisällä
the-house.gen inside
‘inside the house’ (Karlsson 1999: 97)

c. Saksalaisten täytyy lähteä.

the-Germans.gen must leave
‘The Germans must leave.’ (Karlsson 1999: 98)

The genitives exemplified so far in this chapter have all been synthetic; that is,
in all of them, genitive case is marked by morphological means. As was noted
in Chapter 2, case relations can also be marked analytically; thus, for example,
the German synthetic genitive examples (2b) and (2d) could also be rendered
by means of an analytic construction, as in (2bʹ) and (2dʹ).3

(2) b.ʹ der Sänger von dem Lied

the singer of the.dat song
‘the singer of the song’

d.ʹ die Bemessung von der Entschädigung

the calculation of the.dat compensation
‘the calculation of the compensation’

In Chapter 2 it was also noted that, for the purposes of the present investiga-
tion, such non-synthetic marking is not treated as case morphology although

3 The dative markers in (2bʹ) and (2dʹ) are a consequence of the preposition von ‘of’ taking a
dative complement; the nature of the von-construction is returned to in Section 3.4.
36 chapter 3

such constructions are semantically equivalent to their synthetic counterparts.

Indeed, Lander (2009: 584) states that there may be reasons for considering
“analytic genitives” to be “case expressions”.
In view of this investigation’s focus on the competition between the mor-
phological genitive and analytic alternatives in the Germanic languages, it is
worth noting that such competition also exists elsewhere. In colloquial Modern
Greek, in which, as in German, the genitive is more frequent in formal writing
than speech, a prepositional phrase is often used instead of the morphologi-
cal genitive, generally with nouns denoting an inanimate referent (Stephany &
Christofidou 2008: 9).
Blake (2001: 157–158) posits a case hierarchy (7) according to which, “[i]f a
language has a case listed on the hierarchy, it will usually have at least one case
listed from each position to the left”. For example, modern standard German
has a dative case, and it also has nominative, accusative and genitive cases.
However, the German dialects fit less easily into the hierarchy, lacking a geni-
tive case but retaining a dative, and in some dialects combining the accusative
and dative (Section 6.4.2); furthermore, the development of standard German
towards a nominative-accusative-dative system as posited by some scholars
(e.g. Gallmann 1990: 287–288; Section 6.1), likewise does not conform to the
hierarchy. These issues would be overcome should one assume that the von-
construction (defined presently in Section 3.4) functions as a “genitive” (see,
e.g., de Wit [1997: 4], who analyses the cognate Dutch van-construction in this
way).4

(7)5 nominative accusative/ genitive dative locative ablative/ others

ergative instrumental

3.3 The Genitive in the Germanic Languages

3.3.1 Introduction
Alongside concordial case morphology as a whole, the genitive case has been
affected by deflection in the Germanic languages; it was already possible by
the 1950s to look back at “ein lebhaftes Interesse” [‘a lively interest’] which
the topic had attracted (Wellander 1956: 156). With the exception of Icelandic
and Faroese, there are two diachronic developments affecting the genitive case

4 In de Wit (e.g. 1997: 4), van-phrases such as these are termed “[p]ostnominal genitive phrases”.
5 Originally (68) in Blake (2001: 157).
the genitive case 37

which are common to the Germanic languages: the masculine/neuter gen-

itive -s marker became reused as an invariant marker of possession, and a
periphrastic construction—with a preposition indicating the possessor noun
phrase—emerged as a synonymous competitor of the adnominal genitive. In
English and Mainland Scandinavian, these constructions outlived the concor-
dial genitive; in Dutch and German, in contrast, the concordial genitive con-
tinued to be used, at least in the formal, standard varieties. With one minor
exception (portrayed in Scott [2011a, 2012]), the Dutch concordial genitive was
lost by the first half of the 20th century; its German counterpart survives to
this day. In both Dutch and German, the development of the genitive case
was affected by the codification of those languages. In English and Mainland
Scandinavian, on the other hand, the deflectional developments affecting the
genitive case had concluded before codification began. Thus, a concordial gen-
itive never became part of the standard for those languages, the possessive -s
and prepositional constructions taking on the role of “genitive” described in
those languages’ early grammars. In this section, the diachronic development
of the genitive case in English and Mainland Scandinavian is sketched out on
the basis of existing research in order to show the effect of deflection, unaf-
fected by the disruptive factor of codification, on genitive morphology.
Throughout the Germanic languages, the most resilient role of the genitive
has proved to be its adnominal use (Wellander 1956: 156). This is the genitive
type that is strongest—and most resistant to competition—in modern stan-
dard German, and the one that was strongest in written Dutch until it was lost
from the language. Indeed, preserved by usage factors, a fragment of the Dutch
adnominal genitive survives in productive use in the modern language. In
both German and Dutch, the adverbal and adjectival uses of the genitive were
already in decline at the time of the earliest grammarians and have remained
weak throughout the history of the standardised languages. This fits with the
“general trend towards restricting the use of the genitive case to its most cen-
tral functions” observed by Allen (2008: 177) for Middle English; it also held in
Swedish. The Old Norse genitive had been used adnominally, adverbally, prepo-
sitionally and adjectivally; genitive objects of verbs were not typical transitive
objects (Toft 2009). Case distinctions were lost during Old and Middle Swedish
(Norde 2001: 250), and the case system broke down during the 15th century
(Platzack 2002: 170). On the nature of case reduction in the Scandinavian lan-
guages, primarily Norwegian, including the rise of the genitive -s marker in late
Old Norse (and paralleled in other Germanic languages), see also Wetås (2008)
and Enger (2010). The use of the genitive with verbs, adjectives and prepositions
was lost (Haugen 1976: 294, Platzack 2002: 170), with the exception of some fos-
silised expressions (Askedal 2002: 176), but remained as a marker of possession.
38 chapter 3

3.3.2 The Genitive in German

Genitive case in German—exemplified in (8), in which the genitive-marking
elements are shown in boldface—is marked within a noun phrase on determin-
ers and adjectives, and on masculine and neuter singular nouns.6 The definite
articles are des (masculine and neuter singular) and der (feminine singular and
all genders plural); the indefinite articles are eines (masculine and neuter singu-
lar) and einer (feminine singular). Other determiners, such as possessives and
demonstratives, follow this pattern: those ending in -es relate to masculine and
neuter singular, and those ending in -er relate to feminine singular and, where
relevant, all genders plural. Adjectives in a genitive noun phrase receive an -(e)n
ending regardless of gender or number. With rare exceptions (see Durrell 2002:
27–29), masculine and neuter nouns heading a genitive noun phrase receive
an -(e)s ending. Syncretism within the German case system means that not all
these markers are exclusively genitive. For instance, the determiners der and
einer (and all other determiners ending in -er) are common to the genitive and
dative singular; thus, einer fleißigen Lehrerin (8b) is—in isolation—ambiguous
between a genitive and a dative interpretation (see, e.g., Flämig [1991: 541],
Ronneberger-Sibold [2010: 90], and, on the dialectal situation, Henzen [1954:
262]). Additionally, adjectival -(e)n is common to the genitive and dative cases,
accusative masculine singular, and plurals in all cases.

(8) a. das Buch des alten Lehrers

the book the.gen.masc.sg old.gen teacher.gen.masc.sg
‘the old male teacher’s book’

b. das Buch einer fleißigen Lehrerin

the book a.gen.fem.sg hard-working.gen teacher.fem
‘a hard-working female teacher’s book’

A structural constraint prohibiting the use of the genitive is formulated in

Duden (2005: 979–980) as the Genitivregel [‘genitive rule’]: only noun phrases
containing a determiner or adjective which can take genitive inflection may
occur in the genitive; the absence of such a determiner or adjective necessitates
the use of one of the other, semantically equivalent constructions (see also
Paul 1919: 327–328; Bondzio 1967: 140; Gallmann 1990: 258, 287; Flämig 1991: 132;

6 The tendencies towards once-only marking encountered in actual usage—in the form of
omission of the final nominal -s on masculine and neuter singular nouns, and the possessive
-s construction—are examined in Chapter 6.
Another random document with
no related content on Scribd:
p. 359] elles seraient originaires d’Es Souk.
Aucun de ces groupes d’Ifor’as, sauf peut-être le dernier, ne
semble se rattacher de bien près à ceux de l’Adr’ar’.
Les Ifor’as de l’Adr’ar’ ne seraient pas de vrais nobles ; leur pays
appartiendrait en droit aux Oulimminden qui l’ont habité longtemps
et à qui les Ifor’as payaient tribut.
Le départ des Oulimminden pour le sud aurait rendu les Ifor’as
maîtres du pays ; tout ceci est peu clair, car, jusqu’en ces dernières
années, jusqu’à l’occupation française, les Ifor’as étaient tributaires
des Kel Ahaggar et leur payaient l’impôt.

Adr’ar’ Tiguirirt. — A une assez grande distance de l’Adr’ar’ des

Ifor’as (125 kilomètres au sud-est) et séparé de lui par un tanezrouft
de roches cristallines, se trouve un autre paquet montagneux,
l’Adr’ar’ Tiguirirt, que le capitaine Pasquier a eu l’occasion de
traverser. Sa constitution paraît analogue à celle de l’Adr’ar’ des
Ifor’as ; les roches cristallines y jouent un grand rôle et l’abondance
du mica y est remarquable, d’après les renseignements que Pasquier
a bien voulu me communiquer.

L’Aïr.

Pour les habitants du pays, l’Aïr[43] est extrêmement vaste ; il

désigne tous les territoires qui dépendent du sultan d’Agadez ; il
comprend tous les terrains de parcours des Kel Gress et s’étend
jusqu’au voisinage de Sokoto.
Les géographes européens ont pris l’habitude de réserver ce nom
à la région montagneuse qui s’étend d’Agadez à Iférouane et c’est
cet usage que nous suivrons. Le mot Aïr (ou Ahir) est employé par
les Arabes et les Touaregs ; il a un synonyme haoussa : Asbin.

Orographie. — L’Aïr est contigu au sud, et probablement à l’est,

à une haute plaine formée de grès et d’argile appartenant
probablement au Crétacé inférieur (argiles et grès du Tegama). A
l’ouest, une région déprimée, la plaine de Talak qui dans sa partie
méridionale contient quelques lambeaux éocènes, lui fournit une
limite assez précise ; vers le nord il se relie au tanezrouft : une
pénéplaine silurienne, avec de rares îlots archéens, commence à une
cinquantaine de kilomètres au sud d’In Azaoua ; les collines basses
qui la recouvrent sont alignées d’ordinaire suivant une direction
méridienne. Entre Assodé et Aoudéras, cette pénéplaine atteint une
altitude voisine de 800 mètres ; elle s’abaisse au voisinage de 500
au nord comme au sud.
Sur cette pénéplaine sont venus se greffer des accidents
volcaniques importants qui donnent à l’Aïr sa physionomie si
spéciale, et justifient presque le nom d’« Alpes Sahariennes » qui lui
a été parfois attribué.
Il y a une assez grande analogie entre l’Aïr et l’Adr’ar’ : tous deux
qui, par leur latitude, devraient être des tanezrouft, forment, grâce à
leur altitude, en plein désert, des sortes de péninsules demi-fertiles ;
ils appartiennent, par leur climat et leur végétation, à la zone
sahélienne.
Dès le 20° de latitude, les deux massifs jumeaux de Tar’azi et de
Zelim annoncent l’Aïr ; l’un et l’autre se dressent, assez à
l’improviste, au milieu de la pénéplaine qu’ils dépassent de 500
mètres. Tous deux contiennent des points d’eau permanents, des
r’edirs analogues à celui d’In Zize, mais leur caractère volcanique
reste à démontrer.
R. Chudeau. — Sahara Soudanais. Pl. VI.

Cliché Posth
11. — KORI TIN TEBOIRAK (SAISON D’HIVERNAGE).
25 km. à l’Est d’Agadez.
 Cliché Posth
12. — UNE CASCADE PRÈS D’AOUDÉRAS.
Après un orage.
R. Chudeau. — Sahara Soudanais. Pl. VII.

Cliché Posth
13. — LE KORI D’AOUDÉRAS, APRÈS L’ORAGE
 Cliché Posth
14. — PRÈS D’AOUDÉRAS (AÏR).
Au premier plan, une repousse de C. thebaïca, simulant un palmier-nain.
Au fond, le massif d’Aoudéras.

Ces deux massifs sont assez accidentés, ils contiennent des

pâturages suffisants pour quelques montures, et sont le repaire
habituel de Tebbous ou d’Azdjer, coupeurs de route, qui enlèvent
souvent quelques chameaux aux caravanes mal gardées. D’après le
kébir d’Iférouane, El Hadj Mohammed, une tente touareg y était
installée presque à demeure, tout au moins ces années dernières, et
prélevait ouvertement un droit de passage sur les marchands de R’at
qui descendaient à Zinder et à Kano.
Les massifs volcaniques sont nombreux et pressés surtout entre
l’oued Sersou et Aoudéras. Beaucoup de sommets dépassent 1 000
mètres ; quelques-uns atteignent 1 400 et le pic majeur du Timgué
s’élève à environ 1 700, dominant de près de 1 000 mètres la vallée
d’Iférouane.
Cette surimposition, à une vieille pénéplaine usée, de massifs
éruptifs jeunes, donne à l’Aïr un aspect surprenant, presque
paradoxal : les vallées sont des vallées de plaine, souvent larges,
parfois bordées de prairies, à pente assez faible ; le travail de
l’érosion y est insignifiant ; leur fond est tapissé de sable, les galets
y sont rares : les sommets qui, d’un seul jet, s’élèvent à 5 ou 600
mètres au-dessus des rivières, font songer à un pays de montagnes
et de ravins : on s’étonne de ne pas voir des lits de torrents
descendus des hauteurs ; on cherche, au pied des escarpements, les
cônes de déjection.

Fig. 20. — L’Adr’ar’ Adesnou, vu de la gorge de l’oued Kadamellet.

Cet aspect singulier est dû à la juxtaposition de deux formations

que l’érosion n’a pas eu le temps de raccorder. Les parois des dômes
sont trop dures, trop abruptes et trop jeunes pour que, dans un pays
où la pluie est rare, le ruissellement ait pu y créer un bassin de
réception. Les orages coulent en nappe sur leurs flancs ; nulle part
les eaux ne se réunissent en masses assez considérables pour
pouvoir remanier sérieusement les parties basses, les lambeaux, non
recouverts par les laves, de la pénéplaine restée presque
horizontale, lambeaux qui forment entre les massifs volcaniques
comme un réseau de couloirs où les caravanes passent aisément.
 L’Aïr fournit d’excellents exemples de ces montagnes créées par
une accumulation de matériaux, accumulation assez rapide pour que
la part de l’érosion dans la production de ces formes de terrain soit
négligeable. Ces montagnes que l’eau n’a pas sculptées, n’ont jamais
formé de chaînes ; elles ont toujours été isolées les unes des autres.
La sécheresse du climat est évidemment pour beaucoup dans le
rôle insignifiant qu’il convient d’attribuer à l’érosion ; mais il faut
aussi faire sa part au facteur géologique : la plupart des masses
éruptives de l’Aïr rentrent dans la catégorie des cumulo-volcans et
des dômes que l’éruption de Giorgios, en 1866, à Santorin, avait
permis à Fouqué d’entrevoir et que, tout récemment, les dernières
éruptions de la Martinique nous ont appris à mieux connaître[44] :
des crêtes, comme l’Adr’ar’ Ohrsane (fig. 74) sont inexplicables par
l’érosion ; on ne peut les comprendre que formées par la
juxtaposition d’aiguilles, analogues à celles de la montagne Pelée :
elles sont le résultat à peu près inchangé d’un phénomène de
construction.
Parfois cependant, dans l’Aïr, les éruptions ont été d’un type plus
banal ; à Aoudéras, de belles coulées de basaltes sont
accompagnées de projections et de bombes volcaniques ; un bassin
de réception a pu se créer dans les cinérites (fig. 73) et la rivière qui
en sort s’est creusé, dans le plateau d’alluvions qui porte le village
d’Aoudéras, un lit qui est en contre-bas de 5 ou 6 mètres.
La plupart des montagnes de l’Aïr sont de couleur foncée comme
celles de l’Adr’ar’ ; ce vernis du désert qui couvre d’une pellicule
noire la plupart des roches, quelle que soit leur couleur propre, est
extrêmement brillant ; à certaines heures, les massifs d’Asbin ont
presque l’éclat métallique. La présence de cette patine foncée n’est
pas constante : l’Ohrsane est rose et jette une teinte claire sur le
paysage ; jusqu’au sud de l’Aïr, les tons de quelques rochers restent
assez variés et ces taches de couleur vive contrastent assez
gaiement avec les montagnes sombres qui forment les masses
principales. Malheureusement, le ciel est souvent brumeux et l’on ne
voit que rarement dans l’Aïr ces jeux de lumière éclatants qui font le
charme de l’Ahaggar. Parfois cependant, après une averse qui a
nettoyé l’atmosphère, le spectacle devient magnifique ; le 22
septembre 1905, du campement de l’oued Kadamellet, au coucher
du soleil, l’Adr’ar’ Adesnou (fig. 20) semblait une masse de bronze
qui se détachait puissamment sur un ciel lie de vin, la teinte sensible
des physiciens ; quelques nuages bleu indigo ajoutaient, à la magie
de la couleur, une nuance imprévue.
La pénéplaine silurienne et archéenne qui sert de socle à l’Aïr se
relie très graduellement au tanezrouft qui lui fait suite au nord ; les
massifs de Timgué (1 700), d’Aguellal, (1 100) d’Akelamellen (1 200)
et d’Agalac (1 400) reposent sur cette partie basse de la pénéplaine.
Entre les puits d’Agalac et d’Aourarène la piste est obligée de
franchir une falaise d’une quarantaine de mètres, orientée est-ouest,
au nord du volcan d’Aggatane ; on accède ainsi à un plateau qui
porte le Bilat (1 400), le Tchemia, le Baghazan (1 400) et le massif
d’Aoudéras (1 400). Sur la route d’Aoudéras à Agadez la descente est
à peu près continue, sauf quelques marches de 3 à 4 mètres, et je
n’avais pas d’abord attribué à ce plateau du sud de l’Aïr une
importance suffisante. Des renseignements nouveaux, dus à
l’amabilité du capitaine Posth, qui a bien voulu mettre à ma
disposition ses levés d’itinéraires (fig. 22) et de nombreux
documents manuscrits, montrent que ce plateau d’Aoudéras est un
trait tout à fait important dans la structure de l’Aïr. La région
montagneuse s’étend beaucoup plus au sud que ne l’indiquent les
cartes les plus récentes ; les Alpes Sahariennes descendent jusqu’à
la latitude d’Agadez ; leur limite est assez nette et peut être tracée
avec précision ; le rebord méridional de ce plateau est indiqué non
pas par une falaise continue, mais par une série de mamelons, hauts
de 10 à 20 mètres, que l’on peut suivre pendant longtemps au nord
d’une importante vallée qui le sépare du Tegama. Ces premiers
contreforts de l’Aïr avaient été aperçus, de loin, dès 1902, par Cauvin
qui avait escorté, jusqu’à 50 km. à l’est d’Agadez, une forte
caravane.
 Fig. 21. — Aïr. L’Adr’ar’ Timgué ou de l’oued Tidek.
Iférouane est au pied du dernier piton au S.W.

Sur la partie méridionale de ce plateau, qui serait à peu près à

600 mètres, se dressent un certain nombre de massifs montagneux ;
le plus important est le Taraouadji qui contient dans sa partie nord
quelques sommets dont l’altitude varie de 800 à 900 mètres ;
quelques-uns approchent de 1 000 mètres ; la montagne de
Tassamakal et celle de Tsilefin atteignent 800 ou 900 mètres.
Tous ces massifs paraissent en majeure partie granitiques, autant
que l’on en peut juger par les photographies du capitaine Posth et
les échantillons qu’il a rapportés, et qui sont à l’étude au Muséum. Il
est vrai que le pourtour seul des Taraouadji a été parcouru ; d’après
les renseignements des guides, ce massif ne forme pas une masse
compacte ; il est coupé par un grand nombre de vallées souvent
assez larges, du type habituel aux koris de l’Aïr. Les Taraouadji sont
donc très habitables ; en fait, ils ont souvent servi, en cas de
surprise, de refuge aux nomades de la région d’Agadez, et l’on
comprend le peu d’empressement que les gens du pays aient eu à
nous faire connaître leur citadelle.
 RÉGION MÉRIDIONALE
DE L’AÏR

Fig. 22. — Région méridionale de l’Aïr, d’après les itinéraires et les renseignements
du capitaine Posth.

Hydrographie. — Ce haut massif de l’Aïr qui, de l’Ohrsane au

Kori d’Idelioua, se développe sur environ 260 kilomètres avec une
largeur qui en atteint parfois 75, donne naissance à de nombreuses
rivières qui, toutes, coulent trois ou quatre fois par an. Les Haoussa
les appellent des koris[45], le nom est peut-être bon à conserver :
elles sont beaucoup plus vivantes que les oueds sahariens ; dans
quelques-uns de ces koris la végétation est presque forestière, au
sens qu’a ce mot en Europe ; le plus souvent, le kori est couvert
d’un tapis de graminées avec quelques arbres isolés. Dans l’Adr’ar’,
comme dans l’Aïr, la formation végétale qui domine dans les vallées
se rattache à la savane ou à la brousse à mimosées ; les hauteurs
dénudées appartiennent au type saharien ; mais à côté de cette
ressemblance générale il y a des différences nombreuses ; les larges
plaines d’alluvions argileuses sur lesquelles s’épandent en couches
minces les eaux de l’Adr’ar’, n’ont pas d’équivalents dans l’Aïr ; les
vallées sont plus étroites, plus resserrées entre les massifs
montagneux ; leur fond est occupé par des arènes granitiques ou du
sable assez grossier ; il y a parfois des galets ; assez fréquemment
on peut distinguer un lit mineur avec des berges de quelques
décimètres et qui se continue sur tout le parcours de la rivière ; la
pente des vallées, plus forte dans l’Aïr que dans l’Adr’ar’, explique
suffisamment ces divergences.
Les crues doivent être parfois très violentes : dans le haut
Teloua, qui est encaissé, des graminées et des branches charriées
par la crue étaient accrochées à des arbres à 3 mètres du sol.
J’ai vu l’Ir’azar couler à Iférouane, le 7 octobre 1905 ; il avait plu
dans la nuit sur le Timgué ; au matin il y avait dans le ruisseau,
large d’une dizaine de mètres, 0 m. 25 d’eau qui coulait rapidement :
on en entendait le bruit à 100 mètres ; à neuf heures et demie, il
restait quelques flaques isolées, qui disparurent avant midi. L’Adrar
Timgué est imperméable, d’où le peu de durée de la crue.
A Aoudéras, la montagne est formée de coulées de basalte et de
cinérites ; aussi, quand il a plu, le très mince filet d’eau courante qui
passe au pied du village persiste plus longtemps ; nous l’avons vu le
23 octobre ; il avait à peine 1 mètre de large et 2 ou 3 centimètres
de profondeur : les habitants d’Aoudéras comptaient qu’il ne serait à
sec qu’une quinzaine de jours après la tornade. Il y aurait même sur
le Baghazam un ruisseau presque permanent.
 Fig. 23. — Aïr. Extrémité nord du massif d’Akelamellen. — Du puits d’Agalac.

La plupart des belles photographies que Posth a rapportées d’Aïr

ont été prises après des orages. Les ruisseaux et les cascades
qu’elles figurent, donnent du pays une représentation qui n’est que
très accidentellement exacte.
Je n’ai pas vu de r’edir, mais il y en a sûrement dans la
montagne ; ils sont d’accès difficile et les puits sont assez fréquents,
assez peu profonds (18 mètres au plus) pour que l’on puisse
négliger les autres ressources.
A part deux ou trois koris, connus seulement par
renseignements, K. de Tafidet, de Ténéré, qui se dirigent vers l’est et
appartiennent au bassin de Bilma, toutes les eaux de l’Aïr
aboutissent, théoriquement au moins, au bassin du Niger ; il est
douteux qu’une seule goutte d’eau tombée en Asbin arrive aussi loin,
mais l’ancien cours de l’Ir’azar d’Agadez est jalonné par une série de
puits peu profonds, dont quelques-uns sont comblés aujourd’hui ;
celui d’Assaouas (10 m.), à 50 kilomètres d’Agadez, est encore bien
vivant ; à Teguidda n’Adrar, il y a plusieurs mares, alimentées par
des sources qui donnent naissance à de courts ruisseaux. Les puits
suivants : Sekkaret (7 à 8 m.), Tamat Tédret (2 m.), Tamayeur (1-2
m.), Inerider (4 m.), Manetass (4-6 m.), Gessao (1-2 m.) se
succèdent assez régulièrement vers l’ouest. A Tenekart, le fleuve, qui
a pris le nom d’Azaouak, est bien marqué ; la vallée, nettement
encaissée, a 6 ou 7 kilomètres de large. A ce point, l’Azaouak change
de direction et va tout droit vers le sud en passant par Filingué,
Sandiré ; dans cette dernière partie de son cours, il devient le Dallol
Bosso, affluent du Niger[46].
R. Chudeau. — Sahara Soudanais. Pl. VIII.

Cliché Posth
15. — CASES DU VILLAGE D’AGUELLAL (AÏR)
 Cliché Posth
16. — LE MASSIF ET LE VILLAGE D’AOUDÉRAS (AÏR).
R. Chudeau. — Sahara Soudanais. Pl. IX.

Cliché Posth
17. — LE PUITS DE TINCHAMANE, A AGADEZ.
Remarquer l’outre à manche et la double corde.
 Cliché Posth
18. — LES “ DOUM ” (CUCIFERA THEBAÏCA DEL.) DANS UN KORI D’AÏR.

De Tenekart à Assaouas tous les puits mentionnés se trouvent

sur la route de Gao à Agadez, route qui a été suivie par
d’importantes caravanes au temps de la splendeur de l’empire
Sonr’aï.
Pratiquement la masse principale des eaux s’arrête beaucoup
plus près de l’Aïr. Presque toutes les rivières qui prennent naissance
dans la partie méridionale du massif montagneux, dans les
Taraouadji notamment, se dirigent vers le sud et aboutissent à une
région déprimée, allongée de l’est à l’ouest, comprise entre le rebord
du plateau d’Aoudéras et la falaise de Tigueddi ; le long du cours du
kori d’Abrik qui recueille les eaux de cette dépression, se trouvent
plusieurs mares d’hivernage importantes, et des pâturages
permanents, assez fréquentés ; le kori d’Abrik vient rejoindre, à
Assaouas, l’Ir’azar d’Agadez.
Mais la grande majorité des rivières de l’Aïr le traversent de l’est
à l’ouest et vont rejoindre l’Ir’azar d’Iférouane qui, à deux jours de
marche des montagnes, s’épand en une vaste plaine, la plaine de
Talak[47], très riche en eau et en pâturages ; d’après les derniers
renseignements que j’ai pu avoir sur ces régions, le Taffassasset
viendrait lui aussi passer dans cette région du Talak.
Cette plaine de Talak serait un vaste cirque entouré de hauteurs,
surtout vers l’est ; l’eau de source y est abondante et excellente. Les
pâturages y sont beaux ; on parle même d’une forêt vierge,
impénétrable par place. En tous cas, cette région de Talak semble
jouer un rôle très important dans la vie des nomades de l’Aïr ; les
villages de la partie montagneuse, simples entrepôts commerciaux,
ne vivent que de produits achetés au dehors ; dans l’Aïr même, les
vallées se dessèchent parfois et les troupeaux ne trouvent pas
toujours à y paître ; dans la plaine de Talak, au contraire, l’élevage
est toujours possible ; les tentes y sont souvent rassemblées.
Ces deux dépressions recueillent, en somme, presque toutes les
eaux des montagnes de l’Aïr, qu’elles limitent très nettement à
l’ouest et au sud ; ce sont parfois, à la saison des pluies, de
véritables fleuves qui coulent pendant quelques heures. Ces fleuves
se réunissent ou plutôt se réunissaient autrefois, lorsqu’il pleuvait au
Sahara, vers l’ouest, au delà des Teguidda, à Tamat Tédret et
contribuaient tous deux à former l’Azaouak.

Les villages. — Comme l’Ahaggar et l’Adr’ar’ des Ifor’as, l’Aïr est

habité par des nomades et par des sédentaires.
Mais la plupart des villages y ont un caractère commercial très
particulier : l’Aïr s’est trouvé de tout temps sur le passage obligé des
caravanes qui, de Tripolitaine ou de l’Ahaggar, vont commercer dans
les territoires plus riches de Kano et de Zinder. Aussi tous les villages
de l’Aïr sont-ils surtout des relais pour les chameaux, des entrepôts
pour les marchandises ; il n’existe de jardins, de cultures, que dans
un très petit nombre d’entre eux. Depuis que la traite a été
supprimée dans les possessions européennes, les grands convois
d’esclaves ont disparu ; n’ayant plus à échanger contre les produits
de la Méditerranée que quelques plumes d’autruche et des peaux de
filali, les chefs noirs ont dû singulièrement restreindre leurs achats et
le commerce est tombé presque à rien. Aussi tous les centres de l’Aïr
sont en pleine décadence, et tous donnent une fâcheuse impression
de ruine et de misère.
La capitale, Agadez, avait 7 000 habitants vers 1850, d’après
l’évaluation de Barth ; c’est aujourd’hui une ville bien déchue :
l’étendue de ses ruines, l’importance de ses cimetières, la hauteur
de son minaret dénotent un centre autrefois florissant. Ce minaret
(Pl. XI) est une pyramide élevée de 20 à 25 mètres, au sommet de
laquelle on accède par un plan incliné en colimaçon ; c’est
certainement une belle construction en terre sèche, qui, d’après la
légende, aurait neuf cent quatre-vingts ans d’âge et daterait d’Almou
Bari, second sultan d’Aïr, à qui les Kel Gress l’auraient offerte. Il ne
reste plus aujourd’hui à Agadez que 200 chefs de cases : tous ont
été réunis un jour, pour un palabre, dans une des pièces du poste
militaire ; il a été facile de les compter. Cela fait tout au plus 1 500
habitants pour la ville.
Une certaine industrie existe dans la ville ; on y fait de fort belles
sparteries, d’un travail soigné : la matière première est fournie par
les palmiers doums dont les feuilles, coupées en lanières fines, sont
bouillies dans l’eau pour en accroître la souplesse. Ces lanières sont
teintes en jaune avec de l’ocre ; en rouge acajou avec des feuilles de
mil ; pour les teindre en noir, on les fait rouir dans certaines mares
dans lesquelles on jette des scories de forge ; le tannin est fourni
par les feuilles. Ces trois couleurs, jointes à la teinte paille des
feuilles séchées, permettent d’obtenir des dessins géométriques
d’une réelle élégance.
 Comme les fabricants de nattes, les bijoutiers d’Agadez ont une
certaine réputation au Soudan ; ils savent ciseler l’argent avec
quelque finesse et le couvrir d’ornements de bon goût. — L’industrie
de la poterie est également développée.
A Agadez même, la culture est insignifiante : les puits sont
éloignés et profonds, celui du poste français (Tinchamane) est à
1 500 mètres d’Agadez et dépasse 21 mètres ; à Agadez même les
puits, dont l’eau est mauvaise, ont un débit insignifiant. Dans ces
conditions, l’irrigation est pénible, presqu’impossible. Mais à
quelques kilomètres au nord, dans la vallée du Téloua, à Alar’sess,
l’eau est à fleur de sol ; les puits à bascule vont chercher l’eau dans
de simples tilmas. La culture y est assez développée, quoique peu
soignée ; les seguias sont mal entretenues et les planches des
potagers, où tout est semé un peu pêle-mêle, n’ont pas la belle
ordonnance des jardins des Oasis ou de l’Ahaggar où la culture est
aussi correcte que chez nos maraîchers parisiens. Cependant, depuis
la décadence du commerce et la gêne qui en résulte pour les
habitants, la culture tend à se développer. C’est un symptôme
heureux qui est assez général au Sahara.
A Alar’sess on cultive fort peu de céréales (mil, maïs, etc.), mais
surtout des légumes qui sont les mêmes que dans l’Ahaggar
(courges, tomates, etc.). Les principales cultures sont l’oignon et la
carotte ; cette dernière plante serait d’introduction récente dans
l’Aïr ; les premières graines auraient été données aux jardiniers par
Foureau (1900) [Jean, l. c., p. 145].
Les animaux domestiques sont peu nombreux ; les chevaux, les
zébus, les moutons existent à peine. Les chèvres sont assez
communes ; beaucoup d’habitants ont des poules, des pintades et
des pigeons ; quelques autruches domestiques sont élevées dans les
cases. Il y a quelques chiens et, en 1905, le sultan possédait un
chat.
Aoudéras (200 habit.) a, au plus, une soixantaine de cases en
terre et en paille, et quelques tentes en sparterie. Le tissage des
nattes y paraît assez développé. Des puits à bascule permettent
d’irriguer quelques jardins ; l’abondance des coulées de basalte au
voisinage, entretient l’humidité des alluvions et, le long de l’oued, il y
a environ 850 dattiers et autant de doums.
Beaucoup de caravanes passent à Aoudéras ; la plupart des
tribus nomades de l’Aïr y ont une maison où elles déposent leurs
provisions de céréales et leurs objets de valeur, confiés à la garde de
quelques bellah.
Assodé est historiquement la capitale de l’Aïr montagneux, la
patrie du chef des Kel Oui, l’anastafidet Yato. Il y a actuellement 69
maisons habitées et peut-être 200 habitants[48]. Gadel y a compté
337 maisons démolies ; la plupart étaient bâties en pierres et de
forme carrée. Un minaret, comparable peut-être autrefois à celui
d’Agadez, est en ruines aujourd’hui ; il aurait été construit il y a un
millier d’années d’après les informations indigènes et se serait
écroulé il y a 4 siècles.
Ceci n’est guère d’accord avec les indications de Barth, qui place
en 1420 la fondation d’Assodé.
Assodé est aujourd’hui en pleine décadence ; l’anastafidet y a
toujours sa demeure officielle, mais il y vient à peine passer
quelques jours par an et réside habituellement dans le Damergou.
Il n’y a pas de cultures à Assodé.
Aguellal mérite à peine d’être cité ; il n’y a que quelques cases et
huttes, des greniers à mil, et une mosquée sans apparence ; les
jardins font défaut. Aguellal est cependant le centre religieux le plus
important de l’Asbin ; le marabout, El Hadj Sliman, y aurait une
centaine d’élèves ; sa bibliothèque, la plus riche du pays, est évaluée
à un millier de volumes. Il appartient à la confrérie des Quâdria, la
seule importante en Aïr, où les Senoussistes ont peu d’influence.
Iférouane, plus connu dans le pays sous le nom d’Ir’azar, est, à
qui vient du nord, le premier village du Soudan ; il y existe quelques
cases carrées en terre, mais les paillottes rondes à toit conique y
dominent déjà ; elles existent seules dans quelques hameaux de
bergers, voisins d’Iférouane, dont ils ne sont que les faubourgs.
Chaque case, qu’elle soit de terre ou de paille, est habituellement
accompagnée de constructions auxiliaires dont la plus fréquente est
une sorte de vérandah, simple toit posé sur quatre pieux à deux
mètres du sol et que l’on retrouve dans tout le Soudan. Toutes les
constructions qui appartiennent à un même chef de familles ont
encloses d’une palissade commune formée le plus souvent de
branches de korunka. Tout cela est bien nègre.
Les cultures d’Iférouane ont un développement moyen ; une
maigre palmeraie (4 250 palmiers) s’y meurt (Pl. XI, phot. 21). Les
céréales, le mil, le blé, un peu d’orge et de maïs, n’y donnent de
récolte que les années humides. Seuls, quelques légumes (tomates,
oignons, concombres, pastèques et menthe, etc.) y sont d’un produit
assuré. Les puits ont une dizaine de mètres de profondeur et l’eau
en est tirée dans des outres à manche, auxquelles sont attelés des
zébus.
Iférouane est surtout un marché, quelque chose comme un
centre d’affaires ; les notables y ont seulement un pied à terre ; ils
n’y viennent qu’en passant, pour assurer le trafic ; leur vraie
résidence est le village de Tintar’odé qui est situé dans la montagne,
à une quinzaine de kilomètres au sud-est, et où sont déposées leurs
réserves. La population stable serait d’une centaine d’habitants.

Fig. 24. — L’Adr’ar’ Timgué, vu d’Iférouane.

 Il existe dans le sud de l’Aïr, dans le Baghazam, des villages peu
importants comme nombre de cases[49] (Elnoulli, Akari, Tassassat)
où se fait un peu de culture maraîchère. Mais la raison d’être de ces
villages est différente ; situés en dehors des routes commerciales, en
des points d’accès difficile, ils servent, en cas d’attaque, de dernier
refuge : la légende raconte que les Kel Aïr y résistèrent, pendant
trois ans, à un puissant sultan du Bornou, qui dut finalement lever le
siège, et regagna à grand’peine ses États. Bien à l’abri des rezzou, ils
servent surtout d’entrepôts, et l’on y trouve toujours des provisions
considérables de mil, de dattes et de sel.
Il est probable que le cas du Baghazam et de Tintar’odé n’est pas
isolé et qu’à côté des villages de commerce, situés sur les routes
caravanières, il existe de nombreuses retraites dans la montagne ;
c’est une nécessité que la peur a inspirée aussi bien à l’Aïr qu’à
l’Adr’ar’ et à l’Ahaggar.
On pourrait allonger cette liste de quelques noms de villages
encore habités ; on trouvera dans Jean une nomenclature plus
complète. Beaucoup de points sont complètement délaissés, qui ont
été importants : d’Agalac, il reste un cimetière et les débris de
quelques cases ; Tin Telloust, où Barth a résidé, est abandonné de
même que Tafidet. Tin Telloust et Tafidet étaient à la lisière orientale
de l’Aïr, sans protection contre les Tebbous ; cette insécurité a causé
leur ruine : même les pasteurs hésitent à profiter des beaux
pâturages de cette partie de l’Aïr.

In Gall. — Bien qu’il soit en dehors de l’Aïr proprement dit, il

convient de mentionner ici, parce qu’il dépend du sultan, le petit
ksar d’In Gall, situé à l’extrémité occidentale de la falaise de
Tigueddi, dans l’Azaouak, à une centaine de kilomètres à l’ouest
d’Agadez. In Gall aurait été fondé au commencement du XVIIIe siècle
par les Icherifan (Posth)[50].
C’est un marché important pour le commerce du sel, qui provient
de Teguidda n’Tecum, situé à 80 kilomètres au N. Les Kel Gress et
les Oulimminden le fréquentent ; il est sur la route directe de Tahoua
à Agadez. L’eau d’In Gall est très bonne, et il y a une assez belle
palmeraie (4 000 palmiers).
On a cru longtemps que les dattes d’In Gall étaient de qualité
inférieure, car, par crainte de pillage, les propriétaires les cueillaient
dès qu’elles commençaient à mûrir ; depuis que la présence de
tirailleurs permet d’attendre la maturité, on a pu s’assurer que les
dattes étaient bonnes. En 1907, le grain ayant manqué, la récolte a
été assez abondante pour nourrir la population pendant trois mois
(Posth).
Cette localité, plus petite qu’Agadez, n’est pas délabrée comme
elle ; les maisons y ont leurs façades et leurs terrasses ornées de
motifs d’architecture d’assez bon goût ; elles voisinent avec de
bonnes paillottes, à toit conique surmonté d’œufs d’autruches, du
type habituel aux villages noirs.
D’après Gadel, il y aurait une centaine de maisons et 300
habitants à In Gall ; 800, d’après Posth. Un poste de tirailleurs y a
été installé pendant quelques mois en 1904 ; il a été rétabli en 1907.
R. Chudeau. — Sahara Soudanais. Pl. X.

Cliché Posth
19. — UN PUITS A BASCULE DANS LA PALMERAIE D’AOUDÉRAS.
Soir d’orage.
 Cliché Posth
20. — UN KORI D’AÏR
R. Chudeau. — Sahara Soudanais. Pl. XI.

Cliché Posth
21. — LA PALMERAIE D’IFÉROUANE (AÏR)
Il n’y a pas de culture sous les dattiers.
 Cliché Posth
22. — LA MOSQUÉE D’AGADEZ.
Le minaret a une vingtaine de mètres.
 Histoire. — L’Aïr est beaucoup plus peuplé que l’Ahaggar et
l’Adr’ar’ des Ifor’as ; sa population est aussi moins homogène, et
l’organisation politique y est très compliquée. Quelques
renseignements historiques (?) sont nécessaires pour l’éclaircir un
peu.
Quelques tribus nomades, Kel Fédé, Kel Gress, Kel Ferouan, Kel
R’arous, Hoggar, etc., sont blanches et appartiennent aux races
méditerranéennes ; mais la plupart des Touaregs de l’Aïr sont des
noirs ou des mûlatres, apparentés de près aux Haoussas qui
auraient été les premiers habitants du pays.
La langue haoussa est très répandue dans tout l’Aïr ; elle est
comprise généralement de tous et paraît employée dans les villages
de préférence au tamachek. On la retrouve dans les noms propres
où « dan », fils, en haoussa, tient la place du « ben » des Arabes ou
du « ag » des Berbères : Yato dan Kasseri est le nom d’un des
principaux chefs du pays, l’anastafidet.
Agadez et In Gall ont été des colonies de Gao au temps de sa
splendeur et la langue sonr’ai y est encore parlée ou tout au moins
comprise (Lt Jean). Quant aux conquêtes bornouannes, dont la
légende a conservé le souvenir, elles paraissent avoir été sans
influence sur le pays.
Avant notre installation, toute récente[51], dans l’Aïr, le sultan
d’Agadez, le serki n’Asbin des Haoussas, commandait, théoriquement
au moins, aux Kel Gress et aux Kel Oui ainsi qu’à une fraction des
Oulimminden.
Le lieutenant Jean a recueilli, avec grand soin, les traditions
historiques des Asbinaoua ; les Kel Gress et les Kel Oui auraient
quitté, vers le VIIIe siècle, le Fezzan devenu trop peuplé (?) ; ils se
seraient installés dans l’Aïr, les premiers à l’ouest, les seconds à l’est
de la route d’Iférouane à Agadez. Les Kel Gress restèrent peu de
temps au contact des Kel Oui ; ils continuèrent leur migration vers le