Coptic Interference in The Syntax of Greek Letters From Egypt Victoria Beatrix Maria Fendel Full Chapter

Coptic Interference in the Syntax of
Greek Letters from Egypt Victoria

Beatrix Maria Fendel
Visit to download the full and correct content document:
https://ebookmass.com/product/coptic-interference-in-the-syntax-of-greek-letters-from
-egypt-victoria-beatrix-maria-fendel/
OUP CORRECTED PROOF – FINAL, 09/08/22, SPi
OX F O R D C L A S SIC A L M O N O G R A P H S
Published under the supervision of a Committee of the
Faculty of Classics in the University of Oxford
The aim of the Oxford Classical Monographs series (which replaces the Oxford
Classical and Philosophical Monographs) is to publish books based on the best theses
on Greek and Latin literature, ancient history, and ancient philosophy examined by
the Faculty Board of Classics.
Coptic Interference in
the Syntax of Greek
Letters from Egypt
V IC T O R IA B E AT R I X
M A R IA F E N D E L
Great Clarendon Street, Oxford, OX2 6DP,

United Kingdom
Oxford University Press is a department of the University of Oxford.
It furthers the University’s objective of excellence in research, scholarship,
and education by publishing worldwide. Oxford is a registered trade mark of
Oxford University Press in the UK and in certain other countries
© Victoria Beatrix Maria Fendel 2022
The moral rights of the author have been asserted
First Edition published in 2022
Impression: 1
All rights reserved. No part of this publication may be reproduced, stored in
a retrieval system, or transmitted, in any form or by any means, without the
prior permission in writing of Oxford University Press, or as expressly permitted
by law, by licence or under terms agreed with the appropriate reprographics
rights organization. Enquiries concerning reproduction outside the scope of the
above should be sent to the Rights Department, Oxford University Press, at the
address above
You must not circulate this work in any other form
and you must impose this same condition on any acquirer
Published in the United States of America by Oxford University Press
198 Madison Avenue, New York, NY 10016, United States of America
British Library Cataloguing in Publication Data
Data available
Library of Congress Control Number: 2022935398
ISBN 978–0–19–286917–3
DOI: 10.1093/oso/9780192869173.001.0001
Printed and bound by
CPI Group (UK) Ltd, Croydon, CR0 4YY
Links to third party websites are provided by Oxford in good faith and
for information only. Oxford disclaims any responsibility for the materials
contained in any third party website referenced in this work.
To my parents Andreas and Brigitte

Acknowledgements
I am deeply grateful to my DPhil supervisor, Andreas Willi, who guided me all

the way through the storms and calms of this study. For his never-ending insist-
ence, his endless patience, his constant encouragement, and the countless hours
he spent reading and discussing with me every single word of this study, I owe
him my greatest thanks. My thanks go also to his opposite number, Gesa Schenke,
who battled through the Coptic half of this study with me and never ceased to be
enthusiastic and thus provide me with reassurance. Furthermore, I owe thanks to
Wolfgang de Melo, who advised me on how to transform a DPhil thesis into a
monograph. For his witty anecdotes, his consistently constructive criticism, and
his enlightening comparisons, I am immensely grateful.
In addition, my college advisors, Amin Benaissa and Helen Kaufmann, the
two directors of graduate studies I worked with during my time at Oxford, Felix
Budelmann and Gregory Hutchinson, and my teaching mentor, Juliane Kerkhecker,
were a constant source of support. They all contributed to creating a stimulating
environment and helped me to overcome all the little hurdles on the way.
Equally, all the staff of the Classics faculty, in particular the Craven committee,
and the Bodleian libraries had their share in making for this wonderful experience.
Furthermore, I owe thanks to Matthias Müller (Basel) for his interminable
willingness to advise me on all questions of Coptic linguistics and for the manu-
scripts of several forthcoming publications of his. Equally, I owe thanks to Willy
Clarysse (Leuven) for providing me with the manuscript of a very insightful art
icle on later Greek formulaic language as well as to Joanne Vera Stolk (Ghent) and
Klaas Bentein (Ghent) for allowing me to read their brilliant forthcoming articles
on post-classical Greek linguistics.
I would not be at Oxford and this study would not exist without the generosity
of Lady Margaret Hall and the Classics faculty, who awarded to me a Clarendon
scholarship. This allowed me to focus fully and exclusively on my academic work.
I will always be deeply indebted to both institutions.
My thanks go further to the two kind native speakers who identified all my
interferences, hypercorrections, and language-internal confusions of patterns, to
Thomas McConnell and Jonathan Griffiths. They analysed my writing at least as
meticulously as I analysed my Greek data.
Finally, too many people have been on this journey with me to name all of
them. For their optimism, their encouragement and their reliability throughout
this journey, I am indebted to Stephanie Johann (Mannheim), Florence
viii Acknowledgements
Becher-Häusermann (Basel), Roxanne Taylor (Oxford/Manchester), Alexander

Fairclough (Cambridge), Alexandre Loktionov (Cambridge), and Matthew Ireland
(Cambridge) as well as my parents, Andreas and Brigitte, and my siblings, Leonard
and Sophie.
Victoria Beatrix Maria Fendel

Oxford
12 April 2021
Contents
List of Abbreviations xi
Sigla xiii
I . SE T T I N G T H E S C E N E
1. Introduction 3
1.1. Bilingual Interference 3
1.2. Egyptian Interference in Greek 4
1.3. Guiding Principles and Outline 13
1.4. Methodology 14
2. Concepts, Contexts, Corpora 23
2.1. Introduction 23
2.2. Evolution and Contact 23
2.3. Bilingualism and Bilinguality in Egypt 42
2.4. Corpus of Texts 62
2.5. Error Typology 77
2.6. Statistical Concepts 94
3. The Basics of Coptic Grammar 99
3.2. Script 100
3.3. Phonology 101
3.4. Lexicon 102
3.5. Verbs and Sentences 103
3.6. Prepositions and Nouns 112
3.7. Clause Connectors 117
3.8. Regionalisms 126
3.9. Loans from Greek 127
3.10. The Mixed Language Debate 129
4. The Grammar of the Corpus (Standard and Variation) 131
4.2. Phonology 131
4.3. Lexicon 133
4.4. Verb Phrases 137
4.5. Adverbial Phrases 176
4.6. Subordinate Clauses 190
4.7. Coordinate Clauses 210
4.8. Influence or Coincidence? 226
4.9. Personal Names 231
x contents
I I . A NA LYSI S
5. Verb Phrases: The Syntax of Arguments 237

5.2. Data 249
5.3. Summary and Conclusion 285
5.4. Excursus: Avoidance Strategies 291
6. Adverbial Phrases: The Syntax of Adjuncts 295
6.2. Data 299
6.4. Avoidance Patterns 318
7. Discourse Markers: The Syntax of Clause-Linkage 323
7.2. Data 337
7.4. Avoidance Patterns 362
8. Formulaic Language: The Syntax of the Epistolary Frame 365
8.2. Standard Patterns and Variants 370
8.3. Data 382
8.5. Personal Names 411
9. Semi-formulaic Phrases: The Syntax of Signposts and Hedges 413
9.2. Standard Patterns, Variants and Variations 418
9.3. Data 431
I I I . C O N T E X T UA L I Z I N G D EV IAT IO N S
10. Summary and Conclusion 445

10.1. Summary: Types of Errors 445
10.2. Conclusion 1: Error Type and Syntactic Domain 448
10.3. Conclusion 2: Contextual Information 455
10.4. Conclusion 3: Language Acquisition 473
10.5. Outlook 482
Appendix: Corpus of Texts 485

Editions of Texts 485
Abbreviations for Letters 486
Bibliography 495
Index of Keywords 529
Index of Passages 532
List of Abbreviations
A Achmimic dialect of Coptic

ACC Greek accusative case
AcI accusative with infinitive
ADV adverbial
AJ archive of Apa John
AN archive of Apa Nepheros
AO archive of the Apiones of Oxyrhynchos
AOR aorist tense (in Coptic, the unmarked form of the aorist tense is referred to)
AP archive of Apa Paieous
ART article (ART.DEF = definite article/ART.INDF = Coptic indefinite article)
ATT attributive
AUX auxiliary verb
B Bohairic dialect of Coptic
CG classical Greek
CINF Coptic causative infinitive
CND Coptic conditional conjugation
CNJ Coptic conjunctive
COP Greek copular verb/Coptic non-verbal copula
CPL complement
CS Coptic circumstantial conversion
DA archive of Dioscoros of Aphrodito
DAT Greek dative case
DcI dative with infinitive
DEF definite
DEM demonstrative pronoun
DOM Coptic direct object marker
f feminine gender
FUT future tense (in Coptic, the unmarked form of the future tense is referred to)
FUTII marked form of the Coptic future tense
FUTIII modal form of the Coptic future tense (optative)
GcI genitive with infinitive
GEN Greek genitive case
IM infinitive marker
IMP imperative
IMPF Coptic imperfect conversion/Greek imperfect tense
INDF indefinite
INF infinitive
ITJ interjection
IUS Coptic jussive
xii List of Abbreviations
L Lyko-Diospolitan dialect of Coptic

LV light verb
LVC light-verb construction
m masculine gender
M Mesokhemic dialect of Coptic
MATT Coptic marker of attribution
MC main clause/independent clause
MG modern Greek
n neutral gender
NcI nominative with infinitive
NOM Greek nominative case
NT New Testament
OPT Greek optative
PCG post-classical Greek
PCS Coptic precursive
PDC predicative
PK archive of the village of Kellis (Greek texts)
PKC archive of the village of Kellis (Coptic texts)
PL plural
PN predicative noun in a support-verb construction
POSS possessive
PR present tense (in Coptic, the unmarked form of the present tense is referred to)
PRII marked form of the Coptic present tense
PRF perfect tense (in Coptic, the unmarked form of the perfect tense is referred to)
PRFII marked form of the Coptic perfect tense
PRN pronoun
PRP preposition
PRT particle
PTC Greek participle
REL relative
S subject
S Sahidic dialect of Coptic
SBJ Greek subjunctive
SC subordinate clause/dependent clause
SG singular
SL source language
STA Coptic stative
SV support verb
SVC support-verb construction
TL target language
VOC Greek vocative case
Sigla
DDbDP Duke Database of Documentary Papyri (http://papyri.info)

Lampe Lampe (1961)
LSJ Liddell, Scott, and Jones 1996 (http://stephanus.tlg.uci.edu/lsj/)
OED The Oxford English Dictionary (https://www.oed.com/)
TLG Thesaurus Linguae Graecae (http://stephanus.tlg.uci.edu)
Wb Erman and Grapow (1926)
PART I
SET T ING T H E S CE NE
1
Introduction
1.1 Bilingual Interference
Bilingual interference refers to a language user inadvertently drawing on two

rather than one language when putting their ideas into words (Adams 2003
p. 426; Grosjean 2001 p. 6; Myers-Scotton 2006 p. 242). The resultant wording
may be incorrect in the surface-level language, by comparison with the standard
grammar of this language. [1a] is a modern example:
[1a] the man on the photograph.
The preposition ‘on’ is incorrect in English. Supposing that the system of another
language interfered with English in [1a], we may consider [1b] and [1c]:
[1b] l’homme sur la photo.

[1c] der Mann auf dem Photo.
Both French sur and German auf could account for English on in [1a]. While
the surface-level language in [1a] is English, the preposition on is conceptually
incorrect in English. English, unlike French and German, profiles everything
represented by a photograph as contained in the photograph rather than put on
top of the photograph (Lakoff & Johnson 1980 image-schemata). Bilingual inter-
ference is the interaction of two languages in an individual’s linguistic output,
their idiolect. In being idiolectal, interference differs from other contact phenom-
ena such as borrowing and convergence (see Chapter 2).
The present study investigates the surface-level language Greek and the inter-
fering language Egyptian in fourth- to mid-seventh-century Egypt. For most of
its history, Egyptian was written with more than one writing system. In the
Ptolemaic and Roman periods, the diachronic stages of the language are trad
itionally referred to by the name of the writing system used in everyday contexts,
that is Demotic in the Ptolemaic and early Roman periods and Coptic in the later
Roman and early Byzantine periods (Houston et al. 2003; von Lieven & Lippert
2016). In the early Byzantine period, Greek and Coptic were used alongside each
other as papyrological evidence suggests: Archaeologically, Greek and Coptic
documents dating from this period were found in Egypt with documents written
Coptic Interference in the Syntax of Greek Letters from Egypt. Victoria Beatrix Maria Fendel, Oxford University Press.
© Victoria Beatrix Maria Fendel 2022. DOI: 10.1093/oso/9780192869173.003.0001
4 Coptic Interference in the Syntax of Greek Letters
in both languages often discovered together. Linguistically, there are bilingual

documents (e.g. PKC 22) as well as documents that contain Greek passages in
Coptic script or vice versa. Historically, studies of early Byzantine Egypt such as
Bagnall (2007a) confirm a situation of bilingualism. [2] reflects people’s aware-
ness of this:
[2] PKC 19, ll. 13–14 (Kellis, 4th c. ad)

ⲙⲉⲗⲉⲧⲉ ⲛ-ⲛ[ⲉⲕ]-ⲯⲁⲗⲙⲟⲥ ⲉⲓⲧⲉ ⲛ-ⲟⲩⲓⲁⲛⲓⲛ ⲉⲓⲧⲉ ⲛ-ⲣⲙⲛ-ⲕⲏⲙⲉ ϩⲟⲟⲩⲛ <ⲛⲓⲙ>
melete n-n[ek]-psalmos eite n-ouianin eite n-rmn-kēme hooun <nim>
take.care.IMP DOM-POSS.PL.2sg-psalm either in.PRP-Greek or in.PRP-Egyptian
day <every>
Study1 your psalms <every> day either2 in Greek or in Egyptian!
The coexistence of Greek and Egyptian for more than a millennium seems to have
resulted in a situation of stable bilingualism by the fourth century (cf. Thomason
2001 p. 23) as opposed to language shift (Adams 2003 pp. 367–80; Hamers &
Blanc 2000 p. 22; Meakins 2013; Thomason 2001 pp. 21–5 and 66–76) or even
language death (Romaine 2010).3 An example of the latter is the disappearance of
minority languages in the Roman empire (J. Clackson 2012; Wilson 2012).
1.2 Egyptian Interference in Greek
1.2.1 Approaches
Past research into Egyptian interference in Greek (i) has focused on periods
before the emergence of the Coptic alphabet, (ii) has taken the form of smaller-
scale studies, (iii) has been conducted without a cognitive-linguistic component,
and (iv) has pursued non-linguistic, such as historical or archaeological, goals for
which the linguistic analysis was a steppingstone only.
The corpora that received scholarly attention are the agoranomos-contracts
(Papathomas 2007; Vierros 2012a) and the Zenon archive (Evans 2007, 2010a,
2010b, 2012a, 2012b, 2015) dating from the Ptolemaic period as well as the
Narmouthis ostraca dating from the Roman period (Bagnall 2007b; Leiwo 2003;
Rutherford 2010). One may add Clarysse’s (1993) selection of texts, which are
written with an Egyptian rush, and his selective study of the language of the
archive of Kleon and Theodoros (Clarysse 2010a), both dating from the Ptolemaic
1 Greek μελετάω (Förster 2002 p. 510). 2 Greek εἴτε . . . εἴτε (Müller 2017 pp. 288–97).
3 See Section 2.3 (domain stability).
Introduction 5
period. Additionally, Gignac (2013) hints at several structures where he suspects

bilingual interference in a selective study of Ptolemaic and Roman Greek.
Other than these, the same examples, such as the lack of case inflection in
personal names (Fewster 2002 pp. 238–9; Torallas Tovar 2010 p. 262; Vierros 2003
p. 16, 2007 pp. 720–1) and confusion over case endings (e.g. Fewster 2002 p. 235;
Stolk 2015 pp. 22–4), are requoted (see further Section 4.9). Alternatively, a small
number of illustrative examples is discussed selectively, as by Mussies (1968),
Clackson (2010), and Torallas Tovar (2010). Or, descriptive statements without a
full-fledged linguistic analysis of the data are provided (Fournet 2009 p. 442;
Luiselli 2008 p. 715; MacCoull 1987 pp. 311–12; Richter 2014 p. 137).
The treatment of the data depends on a study’s goals. Studies investigating
large-scale historical developments have tended to focus on quantitative statistical
data, such as the number of texts written in Greek and Egyptian found in one
place. They explore the content of the relevant texts and its relation to other
sources, such as literary works.4 In that, the emphasis is often on single illustrative
passages and their meaning. An example is the description of Cleopatra’s language
skills with reference to Plutarch, Life of Anthony 27.3–4 (Thompson 1994 p. 74).
By contrast, studies concerned with individuals’ social context are interested in
the type of the Greek and Egyptian texts. For example, texts may be official con-
tracts as opposed to private records. Based on this data, inferences about language
usage in several domains of life can be made. Additionally, specific passages such
as formulae referring to a person’s literacy (Choat & Yuen-Collingridge 2009;
Kraus 2000) or the clause ὅτι οὐκ ἐπίστ̣αμαι ἑλληνίζειν ‘I do not know how to speak
in Greek’ / ‘behave in a Greek way’ in P. Col. Zenon II 66.21 (= P. Col. 4 66.21)
render insights (Rochette 1996).
Finally, studies exploring the overall linguistic situation in Egypt primarily rely
on qualitative/philological analysis of the data. While a text may be written in
Greek on the surface, it may contain clear indications of an insecure writer.
1.2.2 Multiple Influences
In her contribution to A Companion to the Ancient Greek Language, Torallas

Tovar (2010 p. 259) writes:
[3] Egyptians learning Greek often reached a high level of proficiency and thus
many documents produced by them cannot be distinguished from documents
produced by native speakers of Greek since one cannot identify divergences from
the correct language.
4 Henkelman et al. (2011 p. 465) argue not to take literary works at face value.
Conversely, in his description of the early Byzantine period in The Oxford

Handbook of Papyrology, Fournet (2009 p. 444) terms P. Ross. Georg. IV app.
pp. 99–105 ‘a letter full of Copticisms’ (similarly MacCoull 1987 pp. 311–12).
The letter, in fact, shows the impact of different factors, including but not
limited to bilingual interference, on someone’s writing:
[4] P. Ross. Georg 4 app. pp. 99–105 (Aphrodites Kome, Antaiopolites,

ad 619–29)
† εἶδεν ὁ ἀγαθός μο(υ) δεσπότης ὅτι πολὰ κόπον ἔπαθα κ[αὶ(?) ̣]μ̣[ ̣ ̣ ̣ ̣ ̣ ̣ α]
υτα ἀλλὰ
My good master (may) be aware that I have suffered a lot from beating and . . . but
καὶ τοὺς Πέρσο\υ/ς ἦλθεν ἐν Τηνί ἠει5 καὶ ἐξέν\ε/κέν μοι εἰς τω Φοσᾶτον καὶ
ἐβασάνισέν μοι
also the Persians, (they) came, stayed in Tinis and (then) carried me away to
Fustat and (they) interrogated me
ἀπὸ ὠξιτήου καὶ μαρμάρον εἰς τὼ στόμα καὶ εἰς τὴν ῥῆναν καὶ ἀπέθανα ὁς ἕναν ἐκ
with vinegar and stones in my mouth and nose and I (would have been) dead like
one from
το῀ ν μνιμίον καὶ μετὰ τρῖς ὅρας ἦλθέν μοι πνόην καὶ ηὗρον αὐτο[ὺ]ς ἀπίλθασιν καὶ
καίασέν μοι6
the tombs but after three hours (fresh) air reached me and I found them, (who)
had gone away and left me
ἐκ βορᾶ τῆς Λυτοῦς καὶ ὁ θεὸς ἐβοήθισέν μοι ἐξήλισα εἰς τὰς χύρας αὐτο῀ ν καὶ ἦλθα
north of Letopolis. Because God helped me, I (managed to) escape from their
hands and I went
εἰς Ἀρσενοΐτην καὶ ἰδοὺ ἀειτισώμην. κιμουμένω <μου> καὶ τὰ πεδία μο(υ) ἔλαβεν
εἱ Πέρσις εἰς
to the Arsinoite nome, but behold I was unwell. While I was sleeping, the Persians
took even my children
ἐμʼ. ἔτη [ἐ]γὼ μώνον κιμνὸς ἦλθα ἐνταῦθα. παρακαλ[ο῀ ] τῷ ἀγαθον μου δεσπότῃ.
away from me. Furthermore, I arrived there being alone and naked. I appeal to my
good master.
5 CG ἦ(ν) ‘he was’ (Jernstedt 1927).

6 καίασέν μοι stands for καὶ εἴασάν με. The third-person singular is used as a default form. μοί for
μέ appears repeatedly in the letter.
Introduction 7
The following phonetic spellings appear in the letter:
• iotacism (οι/η/ει/υ), e.g. l. 4 μνίμιον for μνημείων;

• issues surrounding vowel length (ο/ω), e.g. l. 4 το῀ ν μνιμίον for τῶν μνημείων;
• <ε> for <αι>, e.g. l. 6 τὰ πεδία for τὰ παιδία;
• issues with gemination, e.g. l. 5 βορᾶ for βορρᾶ;
• <σ> for <ζ>, e.g. l. 6 ἀειτισώμην for ἀηδιζόμην;
• issue with weak final <ν>, e.g. l. 4 πνόην for πνόη;
• <κ> for <γ>, e.g. l. 7 κιμνός for γυμνός;
• <τ> for <δ>, e.g. l. 3 ὠξιτήου for ὀξιδίου.
The letter exhibits various features of post-classical morphology and syntax.

Morphologically, we see the conflation of the thematic and athematic aorist-
patterns, e.g. l. 1 ἔπαθα for ἔπαθον, as well as the regularization of masculine
second-declension nouns in <ης>, e.g. l. 2 τοὺς Πέρσο\υ/ς for τοὺς Πέρσας, and
the accusative of the third declension with an additional <ν>, e.g. l. 3 ἕναν for ἕνα.
Furthermore, the diminutive ending -ιον appears to lack semantic value
(Horrocks 2014 pp. 175–6), e.g. l. 3 ὠξιτήου (ὀξιδίου) from ὀξίς, ὀξίδος, ἡ. This is
a development relevant from the Roman period onwards and essentially serving
to regularize declensional paradigms.
Syntactically, we observe the use of the present indicative for the future indica-
tive/present subjunctive (Lucas 2014; Markopoulos 2009), e.g. l. 1 εἶδεν for εἰδῇ,
as well as the unclassical choice of prepositions regarding the indication of loca-
tion vs direction and source vs instrument (Luraghi 2003 pp. 122–3, 322, 332),
e.g. l. 2 ἐν Τηνί for εἰς Τήνιν, l. 3 ἀπὸ ὠξιτήου for διὰ ὠξιτήου, ἐν ὠξιτήῳ, or
ὠξιτήῳ. Furthermore, the focus particle ἰδού appears, e.g. l. 6 ἰδοὺ ἀειτισώμην
(see Chapter 7).
Colloquial syntax surfaces in:
• the use of the third-person singular of verbs as the default,7 e.g. l. 2 ἦλθεν,
ἐξένεκεν, ἐβασάνισεν for ἦλθον, ἐξένεκαν, ἐβασάνισαν;
• asyndetic sentence connection, e.g. l. 4 ηὗρον αὐτο[ὺ]ς ἀπίλθασιν καὶ καίασέν
μοι ‘I found them to have gone away and left me’;
• the ‘and’-style throughout the letter.
This leaves three unclassical passages, which Jernstedt related to Coptic idioms:
7 One may argue for interchange between <ε> and <ο> in final unaccented syllables (Horrocks
2014 pp. 112, 118; Leiwo 2003 p. 8). This does not apply to ἠει ‘to be’ in l. 2 and is difficult to uphold
where an enclitic pronoun follows the verb (ἐξέν\ε/κέν μοι, ἐβασάνισέν μοι, καίασέν μοι).
[5a] l. 1 πολὰ κόπον ἔπαθα

literally: I frequently suffered from beating.
ⲁ-ⲓ-ϣⲡ ϩⲁϩ ⲛ-ϩⲓⲥⲉ
a-i-šp hah n-hise
PRF-1sg-receive much of.PRP-beating
I suffered from much beating.
Since πολά (CG πολλά) is neuter and κόπον is masculine, a construal as ‘much
beating’ (πολλὰ κόπων) seems less likely than construing πολλά as an adverbial
accusative. This represents a slight modification of the supposedly underlying
Coptic idiom. Such modifications are common when idioms are translated word
by word.8
[5b] ll. 3–4 ἀπέθανα ὁς ἕναν ἐκ το῀ ν μνιμίον

literally: I was dead like one from the tombs.
ⲁ-ⲓ-ⲙⲟⲩ ⲛ-ⲑⲉ ⲛ-ⲟⲩ-ⲉⲃⲟⲗ ϩⲛ-ⲛⲉ-ⲙϩⲁⲁⲩ
a-i-mou n-the n-ou-ebol hn-ne-mhaau
PRF-1sg-die in.PRP-manner of.PRP-ART.INDF.SG-out of-ART.
INDF.PL-tomb
I was dead like one from the tombs.
Jernstedt (1927 p. 103) considers ὁς (ὡς) equivalent to MG σάν (ὡσάν), a preposition

meaning ‘like’ / ‘as though’. This preposition calls for a complement in the accusative.
The accusative ἕναν is analogous to other third-declension accusatives with an
additional -ν in this period. Alternatively, the Coptic parallel, ⲛ-ⲑⲉ ⲛ- n-the n-,
may account for the accusative after ὡς. While ὡς introduces an elliptic subordinate
clause, ⲛ-ⲑⲉ ⲛ- n-the n- is a preposition. Greek prepositions call for oblique
cases. Without additional evidence, it is impossible to decide which explanation
accounts for [5b]. This is a situation frequently arising in the analysis.
[5c] l. 5 ἐξήλισα εἰς τὰς χύρας αὐτο῀ ν

literally: I escaped towards their hands.
ⲁ-ⲓ-ⲣ-ⲃⲟⲗ ⲉ-ⲛⲉⲩ-ϭⲓϫ
a-i-r-bol e-neu-kjidj
PRF-1sg-do-outside of-POSS.PL.3pl-hand
I escaped from their hands.
8 Phraseological transfer is feasible but impossible to prove here (see Chapters 8 and 9).
Introduction 9
Multivalent Coptic ⲉ- e- may explain the choice of εἰς, which does not suit the
semantic context (see Chapter 6).
On balance, Fournet’s judgement, ‘full of Copticisms’, oversimplifies the situ
ation (cf. Evans 2012a p. 122). The letter is a prime example of the multitude of
influences that impacted on writers’ texts. Bilingual interference only shines
through in three idiomatic passages. Similarly, several instances discussed in
Mussies (1968) can be explained as features of post-classical Greek, as colloquial-
isms, or as mistakes resulting from false analogies.
1.2.3 Morphology, Syntax, Phraseology
Instances of bilingual interference exist in all the areas of the Greek language.
Structures that were labelled interference in the past but that do not qualify as
such are discussed in Section 2.2.4. Regarding phonology, it has been argued that
a regional variety of Greek had developed by the second century ad. Studies con-
cerned with the period before the emergence of Coptic are relevant to the present
study since Coptic developed from earlier Demotic and many structures remained
unchanged.
Morphological interferences are rare. Matras (2015 pp. 23–4) argues that the
adoption of inflectional morphology is dispreferred in language contact settings
because of its integral role in the predication grammar. The predication gram-
mar is the primary indicator of the language used. Therefore, the adoption of
inflectional morphology only happens with language shift impending and
entails the renegotiation of identity. By contrast, examples of the adoption of
derivational morphology include ἐν τῷ τόπῳ τῆς ἑστιάσεως rendering ⲙⲁⲛⲟⲩⲱⲙ
manouōm in [6]:
[6] Pachomius, Praecepta 91 (4th c. ad, Theban area) (Lefort 1956 p. 31)
ⲛ-ⲛⲉ-ⲣⲱⲙⲉ ⲙⲟⲟϣⲉ ϩⲛ-ⲧ-ⲥⲟⲟⲩϩⲥ ⲛⲟⲩⲉϣⲛ-ⲣⲁϩⲧⲟⲩ ϩⲓ-ⲧⲟⲗⲟⲙⲱⲛ ⲉⲓⲧⲉ ⲉ-ⲡ-
ⲥⲱⲟⲩϩ ⲉⲓⲧⲉ ⲉ-ⲡ-ⲙⲁⲛⲟⲩⲱⲙ
n-ne-rōme mooše hn-t-soouhs nouešn-rahtou hi-tolomōn eite e-p-sōouh eite
e-p-manouōm
The men usually go in the congregation without a monkish garment and a belt9
either to the gathering or to the dining place.
The Coptic, and generally Semitic (Lipiński 1997 para. 29.21), prefix ⲙⲁⲛ- man-
‘place (of)’ is translated into Greek (Torallas Tovar 2004 p. 172). Naturally, we
9 This is Torallas Tovar’s (2007) interpretation of the terminology.

would expect deverbal nouns in Greek, such as δειπνητήριον ‘dining room’ or just
ἑστία ‘hearth’, ἑστίαμα ‘banquet’.
Syntactic structures reflecting bilingual interference are more plentiful:
(1) When the numeral ‘two’ functions as a quantifier in combination with a

noun, a plural noun is required in Greek, but a singular noun in Coptic with the
singular noun preceding the numeral (Layton 2011 para. 70b; Müller forthcom-
ing chap. 6.3.1). Greek phrases such as Νεφερῶτι δύο (P. Neph. 11.2–3) may thus
reflect Coptic syntax (Kramer et al. 1987 p. 72).
(2) Several Greek verbs require a specific preposition to attach their comple-
ment in the post-classical period. For example, εὔχομαι calls for περί or ὑπέρ ‘about’.
Conversely, Coptic ϣⲗⲏⲗ šlēl ‘to pray’ calls for ⲉϫⲛ- edjn- ‘above’ / ‘on top of ’:
[7] P. Lond. 6 1926, ll. 14 ἐὰν εὔξῃ ἐπάνω μου εἴασιν λαμβάνω
If you pray for me, I will get better.10
The parallel explains the choice of ἐπάνω after εὔχομαι in passages such as [7]
(Torallas Tovar 2010 p. 263).
(3) While Greek operates with a verbal copula (εἶναι), Coptic has a nominal
copula (ⲡⲉ/ⲧⲉ/ⲛⲉ pe/te/ne). This nominal copula is optional in most syntactic
contexts. Clarysse (2010a p. 42) points out that the secondary addition of copular
εἶναι by a writer should hence be interpreted as suggesting a bilingual writer
proofreading their work.
(4a) Vierros identified two interference patterns that are based on word order
and alignment. The first of these is Greek relative clauses under the influence of
Demotic. Writers tend to identify the Greek relative pronoun with the Demotic
relative converter and the subject in the relative clause. This complicates choosing
the correct gender and number of the relative pronoun. Given that the syntax of
relative clauses did not change significantly in Coptic, we can apply these findings
to Coptic. A theoretical example is [8]:
[8] ὁ ἀνὴρ ἥν ὁρᾷς

ⲡ-ⲣⲱⲙⲉ ⲉⲧⲉ- ⲛⲁⲩ ⲉⲣⲟ-ϥ
p-rōme ete- nau ero=f
ART.DEF.SG-man REL.2sg.f see DOM=3sg.m
the man who you (fem.) see.
Assuming that the writer is female, the subject in the Coptic relative clause
would be female. A pronominal subject ‘you (feminine)’ is covert in Coptic such
that ‘who you’ (relative converter + subject) and ‘who’ (relative converter) are both
10 The writer repeats the construction in l. 24.

Introduction 11
represented by ⲉⲧⲉ- ete- (Layton 2011 para. 396). Since the Coptic relative con-
verter and the subject of the relative clause appear in the same slot as the Greek
relative pronoun, these may be equated. The Coptic female subject then triggers
an incorrect female gender in the Greek relative pronoun (Vierros 2008, 2012a).11
(4b) The second pattern identified by Vierros concerns possessive structures.
Coptic employs two main possessive patterns: (i) {possessed} ⲛⲧⲉ- nte- {posses-
sor} and (ii) {possessed} ⲡⲁ-/ⲧⲁ-/ⲛⲁ- pa-/ta-/na- {possessor} (Müller 2022). In
pattern (i), the form of ⲛⲧⲉ- nte- is independent of the form of the possessed. In
pattern (ii), the choice between ⲡⲁ-/ⲧⲁ-/ⲛⲁ- pa-/ta-/na- depends on the form of
the possessed. By contrast, the Greek possessive genitive is independent of the
form of the possessed. This mismatch results in difficulty choosing the correct
number and gender for the article preceding a noun in the possessive genitive.
A theoretical example is [9]:
[9] οἱ παῖδες τῶν πατρός

ⲛ-ϣⲏⲣⲉ ⲛⲁ- ⲡ-ⲉⲓⲱⲧ
n-šēre na- p-eiōt
ART.DEF.PL-child those.of- ART.DEF.SG-father
the father’s children.
In Coptic, ⲛⲁ- na- must be plural because the possessed is plural. In Greek, the
article would have to be singular because the possessor is singular (Vierros 2012a
pp. 195–203).
Several other syntactic structures have been the topic of debate in the context of
bilingual interference, such as confusion between the genitive and dative cases, a
preference for prepositional phrases over bare cases, periphrastic verb forms, such
as ἔχω μαθών in [10] (Bentein 2016), and cleft-sentences, such as ⲛⲓⲙ ⲙ-ⲙⲟⲛⲁⲭⲟⲥ
ⲡⲉ ⲛⲧⲁ-ϥ-ⲃⲱⲕ nim m-monakhos pe nta-f-bōk in [11] (Reintges 2003):
[10] Sophocles, Antigone 1270–2

{Χο.} οἴμ᾽ ὡς ἔοικας ὀψὲ τὴν δίκην ἰδεῖν.
{Κρ.} οἴμοι, / ἔχω μαθὼν δείλαιος.
Chorus: How unfortunate that you appear to have seen justice very late!
Creon: Very unfortunate, I have learned while being wretched.
[11] (Till 1936 p. 11, ll. 9–11)

ⲏ ⲛⲓⲙ ⲙ-ⲙⲟⲛⲁⲭⲟⲥ ⲡⲉ ⲛⲧⲁ-ϥ-ⲃⲱⲕ ϣⲁⲣⲟ-ϥ ⲉⲧⲃⲉ ⲡ-ⲟⲩϫⲁⲓ ⲛ-ⲧⲉϥ-ⲯⲩⲭⲏ
11 See also Mussies (1968) on attractio relativi and Clarysse (1993 p. 198) on the gender of the rela-
tive pronoun.
ē nim m-monakhos pe nta-f-bōk šaro-f etbe p-oudjai n-tef-psukhē

or.PRT which MATT-monk COP.3sg.m PRF.REL-3sg.m-come towards.PRP-3sg.m
about.PRP ART.DEF-health MATT-POSS.SG.3sg.m-psyche
Or which monk was it that approached him regarding the well-
being of
his psyche?
All these structures are too widespread in post-classical Greek to be classified as

interferences. Furthermore, they fit into the diachronic path of the language.
However, we will see below that a particular agglomeration of relevant instances
could indicate a bilingual writer (cf. Aejmelaeus 1982) (see Section 4.8).
Phraseological transfer refers to the importing of a formulaic or idiomatic
expression characteristic of Egyptian into Greek. The resulting structure may be
grammatical or ungrammatical in Greek. In either case, it is unidiomatic.
Relevant instances have been noted by Clarysse (2010a) and Bagnall (2007b)
regarding dating-formulae, by Mussies (1968) regarding formulae in sales con-
tracts and by Jernstedt (1927) as shown above. An example is the invented internal
address, the first sentence of every letter, in [12]:
[12] {ὁ Παῦλος}sender {τῇ Μαρίᾳ}addressee {χαίρειν}greeting

Paul to Mary, best wishes.
The regular order in Greek of this period would be ‘addressee—sender’ rather than
‘sender—addressee’. Conversely, the order common in Coptic internal addresses is
‘sender—addressee’, for syntactic reasons. Therefore, the order ‘sender—addressee’
in [12] contradicts contemporary Greek usage but reflects Coptic patterns of usage.
The choice of this order does not impact syntactic correctness but only idiomaticity.
Finally, lexical loans are discussed in Section 2.5.5 as they do not constitute
instances of interference. However, Torallas Tovar (2004 p. 171), [13a], and Derchain
(1955), [13b], found instances of loan shifts:
[13a] ὅρος ⲧⲟⲟⲩ toou ‘mountain’ and ‘monastery’.

[13b] θάλλος mnh and mnh.t ‘young branch’ and ‘gift’.
Greek ὅρος means ‘mountain’ whereas Coptic ⲧⲟⲟⲩ toou means not only ‘moun-
tain’ but also ‘monastery’. Resulting from this mismatch, Greek ὅρος could adopt
the meaning ‘monastery’. Similarly, Greek θάλλος means ‘young branch’. Yet, based
on the Egyptian near-parallel mnh.t, θάλλος could adopt the meaning ‘gift’.
Lexicalized/idiomatic expressions are naturally difficult to identify. Some
remarks appear in Torallas Tovar (2010). Derchain (2001) considers Herodotus,
Historiae 2.133 ἔς τε τὰ ἕλεα καὶ τὰ ἄλσεα πλανώμεν ‘Let us wander to the wild
olive trees and to the groves’ to be modelled on Egyptian s3b sš.w literally
Introduction 13
‘se promener à travers les marais’ but with the common figurative meaning ‘se
donner du bon temps, batifoler’. Derchain argues for a word-by-word translation
of the idiom s3b (πλανάω) sšw (ἕλεα) and the subsequent addition of καὶ τὰ ἄλσεα
in order to adapt the Egyptian metaphor to the imagery of a Greek huntsman
(chasseur). As the expression does not exist in Demotic, but only in Hieroglyphs,
an intermediate stage of translation may be involved.
1.3 Guiding Principles and Outline
The present study describes the linguistic dynamics of early Byzantine Egypt from
the philological, socio-linguistic, and cognitive-linguistic perspectives. The focus
is on Egyptian (Coptic) impact on Greek syntax in texts that were produced by
presumably bilingual writers. The present study fills two research gaps. The first
surrounds the study of syntactic contact phenomena between Greek and Egyptian
in the early Byzantine period. Work on earlier periods and on other areas of the
language has been done, as shown above. The second concerns the linguistic
description of early Byzantine Greek which arises from the lack of a third volume
(on syntax) to Gignac’s (1976) grammar.
There has been increasing interest in late antique Egypt, from the historical
(e.g. Bagnall 2007a), sociolinguistic (e.g. Bentein 2013, 2016), and linguistic per-
spectives (e.g. Vierros 2012b) as well as in the diachronic development of Greek
into the Middle Ages (e.g. Bortone 2010; Markopoulos 2009). The notion of ‘bad
Greek’ has been refuted convincingly (Bagnall 2007b p. 21; Evans 2012a, 2012b;
Fewster 2002 pp. 233–5; Papathomas 2007) and unclassical structures are instead
viewed as variations holding important information about the texts (Depauw &
Stolk 2014) and their researchers (S. Clackson 2004). However, a full description
of any early Byzantine Greek corpus of texts or a comprehensive grammar of the
language of the period remains a desideratum (Evans & Obbink 2010 p. 11;
Gignac 2013). Chapter 4 provides a description of the key elements of early
Byzantine Greek and their acceptable variants.
The languages and research traditions discussed have usually remained separ
ate. While studies on post-classical Greek are carried out by classicists, theologi-
ans, or papyrologists, studies on Coptic are the domain of Egyptology (and
Coptology). General linguistic concepts are drawn upon by historical linguists
working in the area of Classics or Egyptology. However, recently, there has been
some collaboration between the disciplines of Classics, Papyrology, Egyptology,
and Linguistics (Bentein & Janse 2020; Leiwo et al. forthcoming). Chapter 2 maps
out how the present study joins empirically supported claims and workable meth-
odologies with philological accuracy and the application of general linguistic
notions (e.g. formulaicity and syntactic complexity).
It appears that the contact phenomena documented for a period of stable bilin-
gualism are of a range of types, which include but are not limited to bilingual
interference. Based on these findings, a typology of errors arising from bilingual-
ity is developed (see Chapters 2 and 10). Empirically, the analysis of the data with
bespoke typology of errors shows that not every area of the syntax is affected by
contact phenomena to the same degree and in the same way. In particular, a clear
line emerges between formulaic and non-formulaic contexts. Finally, Chapter 10
adds the cognitive-linguistic perspective to the picture. The contact situation in
early Byzantine Egypt has often been described in socio-linguistic terms only. Yet,
this precludes any explanation of how interferences emerged in the first place.
The following three research questions serve as the golden threads for the
analysis in Part II:
(1) What types of ungrammatical structures occur? What is the relative dis-
tribution of internal confusion and external impact (interference)?
(2) What is the relative distribution of internal confusion and/or interference
in the areas of syntax analysed?
(3) What is the relative distribution of internal confusion and/or interference
across the texts and archives of the corpus? Are there clusters? Can we
hence identify writers that are more likely to have been bilingual?
Chapter 2 comprises a thorough introduction to bilingualism and language

contact. Chapter 2 also introduces the reader to the select corpus of texts and the
types of language errors observed in the corpus. Chapter 3 outlines the basics of
Coptic grammar to the extent that they are relevant for the present study.
Chapter 4 describes the Greek grammar of the corpus. Chapter 5 examines the
area of verbal syntax, that is the argument and participant structures surrounding
predicates. Chapter 6 assesses the area of nominal syntax, that is the syntax of
the adverbial phrase and its embedding in the surrounding sentence structure.
Chapter 7 explores the area of clausal syntax, in that we delve into clause linkage
by means of conjunctions, particles and implicature. Chapters 5–7 are concerned
with non-formulaic contexts. Chapter 8 investigates formulaic contexts, that is
the epistolary frame. Chapter 9 investigates semi-formulaic contexts, that is fixed
expressions used to structure the letter body. Finally, Chapter 10 returns to the
research questions of this study in light of the full investigation of the three areas
of syntax and the three types of contexts and links the results together.
1.4 Methodology
We rely throughout on two premises. Firstly, Greek and Coptic are independent
languages rather than contact varieties. Their description in Chapters 3 and 4
Introduction 15
supports this premise. The premise is relevant for the sampling of data, in that the
Greek and Coptic samples are independent of each other (see Section 1.4.1).
Secondly, the coexistence of two languages (bilingualism) and their use by the
same individual (bilinguality) are cross-linguistic phenomena. This premise allows
us to draw on frameworks and methodologies developed for modern languages.
The differences to be accounted for lie in (i) the lack of native speakers (Adams
2003 p. 3) and (ii) the finite amount of data (Torallas Tovar 2010 pp. 253–4) (see
Section 1.4.2).
1.4.1 Clusters and Interdependencies
Corpus-languages, such as Greek and Coptic, are only accessible through written
sources (Fleischman 2000 pp. 34–5; Langslow 2002 pp. 23–4). Therefore, linguis-
tically, we opt for a corpus- based approach, that is an inductive/empirical
approach. We define a corpus of texts—of documentary rather than literary or
epigraphic data—, collect the relevant data points in the corpus, and analyse the
data philologically and quantitatively.
In order to define the corpus of interest, the notion of ‘archive’ as used in the
field of papyrology is of interest. Papyrus archives are groups of texts that have
been assembled by modern scholars based on the common origin of the texts or
on prosopographical data in the texts indicating that the texts were sent to the
same person or originated from the same community (Clarysse 2010b pp. 48–53;
Jördens 2001; Vandorpe 2009 pp. 226–9). The owner of an archive received or
collected the texts that constitute the archive. Thus, it is not the archive owners’,
but their surroundings’ linguistic ability that we assess by considering an archive.
Occasionally, archives contain several letters by the same person due to long-
term contact. Examples include Philinos and Iason (Evans 2012b pp. 26–9 and
34–9), Zenon, Amyntas, and Hierokles (Evans 2010b pp. 64–6, 2015), and Ktesias,
Hierokles, and Artemidoros (Evans 2007) in the Ptolemaic Zenon archive, as well
as Paul in the early Byzantine Nepheros archive (Kramer et al. 1987 pp. 24–32).12
If an archive contains Greek and Coptic texts,13 we can assume that the archive
originated from a bilingual environment (Fewster 2002 p. 236).
Archives are the incomplete reflection of someone’s social network (Gardner
et al. 1999 p. 6).14 This reflection was always incomplete as someone’s archive only
12 Luiselli (2008 pp. 680–1) finds groups of private and official letters written by the same person in
archives from the Roman and Ptolemaic periods for sixteen individuals including Paul in the
Nepheros archive.
13 Bilingual archives are discussed by Clarysse (2010b), Vandorpe (2009 p. 242 fig. 10.8), Clackson
(2010 pp. 89–99), Fournet (2009 pp. 435–41), Wipszycka (2009 pp. 80–99). See also www.trismegistos.
org (accessed 17 October 2017).
14 See further Section 2.4.4.
documents their incoming correspondence but (usually) does not contain any
outgoing correspondence. A rare exception is the archive of Apollinarios (Vandorpe
2009 pp. 237–8), who kept copies of his own writings.
In order to show clearly which archives papyri belong to, the references to
the letters of an archive are given abbreviations that reflect their belonging to the
archive (e.g. AN for the archive of Apa Nepheros). The reader is referred to
the List of abbreviations for the resolution of these abbreviations. The select
archives are d iscussed in Section 2.4.4.
The advantage of using archives is that we can gauge comparatively well what
kind of relationship existed between the interlocutors involved, that is the writer
and the archive owner. The relevant relationship differs between archives, as for
example someone writing to a desert monk must still consider a certain hierarchy
between the former and themselves, whereas someone writing to a fellow villager
may not have any hierarchy on their mind.
However, using archives complicates statistical analysis. Our sample is not a
simple random sample to start with, but we are looking at cases that are clearly
linked. This is a problem for carrying out statistical testing, such as hypothesis
testing. Furthermore, the texts that have come down to us have undergone sub-
stantial selective procedures that we cannot account for ranging from the chance
of preservation through sorting/throwing out/etc. in antiquity and multiple pre-
modern scholars’ decisions to keep or discard texts to accessibility in modern col-
lections. Thus, in the eyes of a statistician, the data sample is challenging to say
the least. Finally, our sample is very small, and data cannot be multiplied.
Statistically, this raises the issue that standard errors go up as the sample size
goes down.
Nonetheless, what we can do, with caution, is calculate proportions with regard
to the distribution of errors/contact phenomena and test for the difference
between proportions. Furthermore, we can draw on statistical methods devel-
oped for small samples and complex samples (although with the caveat that the
sampling procedure is largely unknown to us) and apply chi-square tests to the
error distributions (see Chapter 2).
In the end, we have a sample of texts the social context of which we can
describe comparatively well and for which we can assume a linguistically speak-
ing mixed environment. This sample of texts is preserved well enough to lend
itself to qualitative/philological analysis. Thus, we rely heavily on qualitative
analysis and support this with quantitative analysis where possible.
Manual data collection and analysis is preferable as the following four examples
taken from the texts of the corpus illustrate:
[14a] AP 1, ll. 19 καὶ τὸν μονάριν Ἡρακλίδην̣ δύ[σα]ντες καὶ ὑβρίσαντες

concerning the monk Heraklides, having bound/plunged and maltreated (him).
Introduction 17
In [14a], orthographic variability impacts on which verbal lemma we catalogue.

Depending on whether we assume iotacism, <υ> for <η>, or not, we may read
either δύω or δέω. Both readings suit the context.
[14b] AJ 7, ll. 7 ἵνα ἀπολύομαι

in order that I may be released.
[14c] AP 2, ll. 13 ἵν<α> καὶ ὑμῖς γνώντες συμβάλλεσθε αὐτῷ
in order that you too, then being aware (of it), may help him.
In [14b] and [14c], orthographic variability affects syntactic correctness. If we

assume a lack of vowel-length distinction between <ο> and <ω> in [14b] and
between <ε> and <η> in [14c] (Clarysse 1993 p. 197), both instances reflect clas-
sical syntax. Alternatively, we may construe the apparent indicatives as modern,
colloquial structures (Hult 1990 pp. 171–2).
[14d] AN 13, ll. 11]πραὺ̣ ἐπιξαμε[ν

we urged . . .
In [14d], orthographic variability is relevant to the voice of the verb. If we read

ἐπείξαμεν, the verb is active; if we read ἐπειξάμην, the verb is middle. The context
is too damaged to decide conclusively.
Besides orthographic variation, frequent abbreviations of words and misspellings
render automatized searches difficult. Often, only the entire context clarifies
which option is more likely. For example, AJ 7 contains several ungrammatical
structures. In light of these, reading an indicative after ἵνα in [14b] seems con-
ceivable. However, AJ 7 also contains various phonetic spellings (Fournet 2019
pp. 50–3) so that interchange of <ο> and <ω> cannot be excluded.
All the Greek texts are quoted in the form in which they appear in the relevant
edition of the text, including spurious apostrophes. Only occasionally a correc-
tion is suggested, primarily with regard to the accentuation of single words. The
version of the text given in the edition is then quoted in a footnote.
1.4.2 Structure, Context, Formation
As we aim to linguistically describe, contextually embed and cognitively explain

the errors that appear in the corpus, three frameworks are relied upon. Lexical-
functional grammar serves to describe philologically the relevant instances; the
sociolinguistic framework serves to embed the relevant instances contextually;
Construction Grammar serves to explain how and why an error occurred. The
three frameworks are briefly described one by one.
Lexical-Functional Grammar (henceforth LFG) is built around three

rinciples, that is variability, universality and monotonicity (Bresnan et al. 2015
p
pp. 56–7).15 The c-structure, that is the constituent or categorical structure, is the
external structure (Bresnan et al. 2015 p. 56). The c-structure is language-specific
(cf. the principle of variability), in that different languages have different surface-
level representations. The f-structure, that is the functional structure, is the internal
structure. The f-structure describes universal grammatical relations (Bresnan et al.
2015 p. 56). Recall the example of ‘in the picture’ vs ‘sur la photo’ / ‘auf dem Bild’,
where we have different surface-level representations of the same underlying
functional structure and concept. The c-structure and the f-structure map onto
each other. When we add up the c-structure and the f-structure attributes of a
given constituent, our description of the constituent becomes more and more
defined. The level of specificity increases monotonically (Bresnan et al. 2015 p. 78).
The ‘argument structure is an interface between the semantics and syntax of
predicators (which we may take to be verbs in the general case)’ (Bresnan et al.
2015 pp. 253–4). It is double-faced, in that we have the representation of the
event structure on the semantic side and the representation of the syntactic sub
categorization on the syntactic side. On the syntactic side, we are interested in
syntactic dependents of a governing head (Bresnan et al. 2015 pp. 253–4), that is
in arguments and adjuncts. On the semantic side, we are interested in core par-
ticipants, that is in essence semantic or thematic roles (Kroeger 2005 p. 54).
Thematic roles are notoriously difficult to define. For example, where do we
draw the line between a Theme, a Stimulus, and a Patient?
[15] (Kroeger 2005 p. 54)

Theme: entity which undergoes a change of location or possession, or whose
location is being specified.
Stimulus: object of perception, cognition, or emotion; entity which is seen,
heard, known, remembered, loved, hated, etc.
Patient: entity which is acted upon, affected, or created; or of which a state or
change of state is predicated.
In prototypical cases, the distinction may be clear, but in reality it is often blurred.
Therefore, Naess (2007) in her book on prototypical transitivity proposed an
alternative way of defining participant roles. She considers the criteria of volition-
ality, instigation, and affectedness for every participant role. Volitionality refers to
the volitional involvement of an entity in an action; instigation refers to an entity’s
instigating role as regards an action; affectedness refers to whether or not an
entity is affected/changed by an action. Applying these three criteria, Naess defines
the following participant roles:
15 A recent up-to-date reassessment of LFG is Dalrymple et al. (2019).

Introduction 19
[16] Participant roles according to Naess (2007 chap. 5).
Volition Instigation Affectedness

+ + – Agent
– – + Patient
+ – + Volitional undergoer (also: experiencer,
recipient, beneficiary)
– + – Force (esp. natural forces, involuntary agents)
– + + Instrument (manipulated by the agent,
impacting on the patient)
+ – – Frustrative (failing to, not being able to)
– – – Neutral (e.g. stimulus, resultative object/
coming into being as a result of the action)
Drawing on three clear-cut diagnostic criteria homogenises the assignment of

participant roles.
For the application of the sociolinguistic framework to classical and post-
classical Greek, we draw on the work by Labov (1963, 1991), Romaine (1982),
Halliday and Matthiessen (2004), Halliday and Webster (2007), and specifically
Willi (2003) and Bentein (2016). Sociolinguistics is a usage-based approach to
language centred on the language users. The approach holds that there is an intri-
cate connection between the social and linguistic realities. The general idea is that
users of language are influenced by their situational surroundings including other
users who they interact with (see in detail Sections 2.2.3 and 2.3.1).
Cognitive approaches focus on language users’ processing of language (Bybee
2010). Unlike LFG, Construction Grammar is a monostratal theory (Goldberg
1995 p. 7). Any formal difference is linked to a functional difference (Goldberg
1995 p. 3). For example, formally, the difference between ‘to be afraid to’ and ‘to
be afraid of ’ is the change in the final particle triggering the subsequent change in
the form of the complement of the particle, that is an infinitive after ‘to’ and an
ing-form after ‘of ’. While ‘to be afraid to’ involves intentionality, ‘to be afraid of ’
does not (Goldberg 1995 p. 3). A construction ‘is posited in the grammar if and
only if something about its form, meaning, or use is not strictly predictable from
other aspects of the grammar’ (Goldberg 1995 p. 13). Constructions exist at all
structural levels, including the clause, the phrase, and the sentence. The change in
meaning between ‘to be afraid of ’ and ‘to be afraid to’ results from the construc-
tional change rather than polysemy of the verb phrase or lexical changes
(Goldberg 1995 p. 18). By contrast, no functional difference exists between
English ‘try to’ and ‘try and’ (Gries & Stefanowitsch 2004 pp. 122–3). Hence, we
cannot posit a construction. Experimental research indicates that language users
have a certain awareness of the close association of, for example, certain verbal
items with specific constructions (Goldberg 1995 p. 36) (see further Section 2.5.1).
1.4.3 Standard, Variation, Deviation
The analysis of structures observed in the corpus of texts proceeds along a triad:
A standard is established, variations of the standard are identified, and deviations
from the standard are pinpointed. Variations are grammatically correct instances
that do not reflect the standard. Deviations are ungrammatical instances that do
not reflect the standard.
The concept of a standard is complex. The seminal work is Haugen (1966), who
proposes that the evolution of a standard results from the interaction of social and
linguistic processes. A social group selects a norm, which is then codified in the lan-
guage and the function of which is elaborated. At this stage, the social group in ques-
tion accepts (or rejects) it. Social groups or social networks thus play an important
role in the spread of new features and the maintenance of old features, depending on
the makeup of the network (Milroy & Milroy 2012 pp. 49–50). Haugen’s model falls
into the category of a standard evolving from acceptance through agreement.
Alternatively, a standard may evolve by prescription through authority (Versteegh
2002 p. 55), a modern example of which is the Académie Française. In this scenario,
the standard is an artificial written ideal that users aim for (Colvin 2009 p. 36) and
which gives rise to a standard ideology. Colvin (2009 p. 43) shows that such a
standard existed already in the classical period. The existence of a standard ideology
and the prestige thus attributed to a standard language can be detrimental to
research efforts as it creates an underlying conceptualization of a bipartite distinction
between correct and corrupted (Hernández Campoy & Schilling 2012 pp. 70–3).
Generally speaking, standard languages do not notate sociolectal (e.g. the
glottal stop in ‘water’), colloquial (e.g. the contractions in ‘isn’t’ and ‘can’t’), or
dialectal features (e.g. the British English north-south isogloss /a/ as in ‘castle’).
Standard languages are likely to surface in non-literary official contexts where the
focus is on the factual transmission or recording of a message.
In Egypt, a standard ideology seems to have existed at least from the Roman
period onwards, as evidenced by the availability of grammar teaching and move-
ments such as Atticism and the Second Sophistic (see Section 2.5.2). Furthermore,
writers occasionally amended their own mistakes. These self-corrections point to
the writers’ awareness of an existing standard language. Scribal corrections are
noted in passing when they appear in ungrammatical or unidiomatic passages
but are not listed systematically in Part II.16
16 Research is ongoing on this topic. Cf. Stolk (Ghent), Scribal Corrections and Language Variation
and Change in Greek Documentary Papyri from Egypt (300 BCE–800 CE); http://research.flw.ugent.be/
Introduction 21
Ideally, we would have a simple random sample taken from all the texts that
existed in fourth- to mid-seventh-century Egypt in order to assess people’s aver-
age linguistic choices, in the sense of a standard as acceptance through agreement
or Hult’s (1990 p. 23) notion of Normalprosa. This is what descriptive grammars
should provide us with. By contrast, prescriptive grammars are influenced by the
standard language paradigm. For early Byzantine Greek and early Coptic, we lack
either kind of detailed grammatical descriptions (S. Clackson 2004; Oréal 1999
pp. 197–302), unlike for the Ptolemaic papyri (Mayser 1906), the New Testament
(Bailey 2009; Blass 1990; Kirk 2012; Moulton 1957; Robertson 1919) and later
Coptic literature. However, there are studies on single phenomena, such as
Markopoulos (2009) on the future or Bentein (2015, 2017) and Hult (1990) on
verbal complementation patterns.
Generally, we need to be careful when drawing on literary sources as a point of
comparison since writers of documentary texts did not aim for literary excellence
but comprehensibility. A multiplicity of factors impacted on their texts including
the diachronic development of the language, the situational register, the writers’
level of education, and their social context (Evans 2012a p. 123). Coptic gram-
mars, in particular, are usually based on literary sources.
For the present study, the standard is defined in the sense of acceptance
through agreement. Three measures are established to distinguish between post-
classical standard Greek in Egypt and deviations from the standard.
(a) Frequency: The more frequent a structure, the more common it is. However,
mere token frequency is not indicative. For instance, we may have a fair number
of tokens, but all of them in the writings of the same person. In this case, the
structure may rather reflect an idiolect than a general trend (Ellis 2003; Stolk 2015
pp. 65–7). The frequency measure is readily applicable only to high-frequency
items (> 10 attestations) that are distributed across a range of writers.
(b) Grammaticality (see further Section 2.5.2): If a pattern appears in classical
Greek, the Ptolemaic papyri, or the New Testament (henceforth NT), it is con
sidered grammatical. The NT is a well-studied corpus of Roman Greek literature.
Despite it being a work of literature, its language is generally regarded as closely
resembling standard Roman Greek (Blass 1990 para. 3; Deissmann 1908 p. 40;
Horrocks 2014 p. 147; Mandēlaras 1973 paras 21–2 and 35; Watt 2013 pp. 21–4).17
However, Gignac (2013 p. 419) emphasizes the bilingual surroundings in which
the NT emerged. On the Coptic side, the situation is more complicated as the
Coptic NT is a translation from Greek (Hasznos 2006, 2012 pp. 74–8; Oréal 1999
en/projects/scribal-corrections-and-language-variation-and-change-greek-documentary-papyri-
egypt-300 (accessed 1 April 2018).
17 Linguistic differences exist between the four gospels, e.g. Luke displays the fewest traces of an
underlying Semitic first language, whereas Matthew allows for contact phenomena from Aramaic and
Latin in his writings (e.g. Jiménez López 2017, 2018).
p. 295). Thus, translation techniques pertain to the shape of the language

(Langslow 2012 pp. 143–5). For either language, the NT has come down to us in a
multiplicity of manuscripts and consequently a multiplicity of versions of the text
(Ehrman 2016 p. 23; Metzger & Ehrman 2005 pp. 50 and 250–99).
The Septuagint is less relevant as it was translated from Hebrew around the
third century bc (Horrocks 2014 pp. 106–8). Hence, there is a time-gap to our
early Byzantine data. Despite being a translated text, the Septuagint seems to
reflect regular diachronic trends. For example, Evans (2015 p. 80) finds this to be
true for the choice of verbal aspect (further Evans 2001).
A pattern is also considered grammatical if it has clearly developed from a pat-
tern attested in one of the three varieties mentioned above reflecting a regular
evolutionary path (see Chapter 2.2.4). Yet, incidentally, the restructuring of many
areas of the language system and the resulting coexistence of several patterns
meant that even native speakers may sometimes have struggled to apply what
seems to have still been the standard pattern. This kind of internal variation pre-
ceding language change is a regular phenomenon (Matras 2009 p. 60; Stolk 2015
pp. 25–6).
(c) Convergence (see further Section 2.2.2): Convergence means that the struc-
tures of two languages fall together and mix rather than there being imposition of
the structures of one language onto the other as there is in interference. When
trends converge, the resulting structure may appear commonly or rarely. In the
former case, it may be a regionalism. In the latter case, a proficient bilingual may
have produced the structure as part of their idiolect.
The above observations are summarized in [17]:
[17] Defining a standard
(a) Frequency (b) Grammaticality (c) Convergence

✓ ✓ – → standard
✓ ✓ ✓ → regionalism
– – – → interference
– ✓ ✓ → idiolect
Structures that have resulted from interference or a proficient individual’s idiolect

are discussed in Chapters 5–9. Structures that represent the standard or constitute
regionalisms are discussed in Section 2.5.6 and Chapter 4.
2
Concepts, Contexts, Corpora
2.1 Introduction
The present chapter is a toolbox of theoretical concepts and frameworks, an

introduction to the chosen corpus of texts and the error typology developed in
Chapters 5–9 and an overview of the statistical methods applied to the error dis-
tributions in Chapters 5–9.
Theoretical frameworks and concepts are linked to the areas of general linguis-
tics, cognitive linguistics, sociolinguistics, and second-language acquisition. The
select corpus of texts is a papyrological corpus and thus the description of the
corpus is embedded in papyrological terminology and considerations. Parametric
and non-parametric statistical approaches to the data are discussed.
2.2 Evolution and Contact
Language change often results from a combination of internal and external fac-
tors that impact on a language variety over a period of time (Traugott & Trousdale
2010 p. 21). Internal factors relate to language-internal evolutionary processes.
External factors relate to the contact with other languages. Internal evolution is
relevant to the present study in order to determine the standard, since any com-
parison of our texts with classical, Ptolemaic, or New Testament Greek would
impute more mistakes to our writers than they actually made.
In theory, we can approach languages from a diachronic or a synchronic
perspective. Practically, the two perspectives can hardly be separated (Aitchison
2012 p. 19). The present study reflects this state of affairs: Although we are going
to take a synchronic approach, the diachronic level of analysis is relevant insofar
as early Byzantine Greek evolved from classical and post-classical Greek and into
medieval and modern Greek. Equally, Coptic developed from earlier Egyptian
(Demotic). Diachronic trends help us distinguish between structures that are due
to internal evolution and those that are foreign to Greek grammar.
Coptic Interference in the Syntax of Greek Letters from Egypt. Victoria Beatrix Maria Fendel, Oxford University Press.
© Victoria Beatrix Maria Fendel 2022. DOI: 10.1093/oso/9780192869173.003.0002
2.2.1 Processes of Language Change: Grammaticalization,

Lexicalization, Pragmaticalization
Three internal processes of language change are grammaticalization, lexicalization,

and pragmaticalization. They all concern items moving from one domain into
another, for example from the lexical into the grammatical or from the grammatical
into the pragmatic domain. The three processes are discussed one by one.
Grammaticalization is the process by which an item moves from having a clear
pragmatic and/or lexical function to being a more functional element in the mor-
phosyntactic structure (Breban 2019; Marchello-Nizia 2006). Grammaticalization
is described as a unidirectional process of language change. Yet the opposite process
(conversion) is attested in some languages (Hopper & Traugott 2003 p. 58).
Grammaticalization is essentially an umbrella term for a range of phonetic,
morphological, syntactic, and semantic changes the occurrence of all or some of
which renders an item more grammatical than before. The observable changes
include phonetic reduction, morphological and syntactic reanalysis, and seman-
tic bleaching. The two major driving forces are reanalysis, a rule change at the
syntagmatic level, and analogy, a generalization of a rule at the paradigmatic level
(Hopper & Traugott 2003 pp. 52–5 and 64). Reanalysis often affects ambiguous
structures (Hopper & Traugott 2003 p. 52). It is the process by which an item is
formally transformed (often in the form of re-segmentation) or/and functionally
transformed resulting in a new functional profile. This is usually due to speaker
economy (Patard 2014 pp. 126–8).
An example at the morphological level is the French inflectional future: classical
Latin cant-a-bo sing-thematic.vowel-future.1sg ‘I will sing’ becomes cant-a-re
habeo sing-thematic.vowel-present.infinitive ‘I will sing’ in later Latin, which in
turn becomes je chant-erai sing-future.1sg in modern French (J. Adams 2013
pp. 822–3; Hopper & Traugott 2003 pp. 52–3). The change from later Latin into
French, i.e. cantare habeo vs je chanterai, involves phonetic reduction and univer-
bation, reanalysis of habeo as a future marker, and semantic bleaching of habeo
(cf. Chapter 4 §21 for Greek). Heine and Kuteva (2002) provide plentiful examples
across languages.
Lexicalization is the process by which an item moves further into the lexical
domain, that is becomes an idiomatic, often semantically non-compositional lex
ical expression. In the process, the mechanisms of reanalysis and analogy affect
the lexicon, rather than the grammar (Hopper & Traugott 2003 p. 58). Lexicalization
entails changes to the semantics and applicability of the item in question, for
example its register- and genre-related applicability. For example, for prendre DP
décision vs rendre DP décision, the meaning of the latter is made narrower through
Concepts, Contexts, Corpora 25
lexicalization with the result that it appears in contexts in which a certain (polit
ical) authority is taking a decision (Fendel 2020).
Opinions are however divided about the exact definition of lexicalization. We
follow Lipka (1981 p. 120): ‘Unter Lexikalisierung verstehe ich die Erscheinung,
daß einmal gebildete komplexe Lexeme bei häufigem Gebrauch dazu tendieren,
eine einzige lexikalische Einheit mit spezifischem Inhalt zu werden. Durch die
Lexikalisierung geht der Syntagmacharakter in mehr oder weniger starkem Maße
verloren’1 (see further Lipka 1992).
Hopper and Traugott (2003 p. 53) offer the example of compounding as in
house+wife. Compounding is a type of reanalysis that usually involves the dis
appearance of the boundary between words and/or morphemes. In housewife, the
element wife has not been reanalysed as a grammatical morpheme (such as
-hood). Yet, wife in housewife is semantically narrower than wife when used on its
own. Moreover, the combination housewife refers to a specific concept that wife
alone could not refer to. The combination has assumed a specific connotation that
wife alone does not have. The meaning of the combination is no longer semantic
ally compositional but has developed further from the originally semantically
compositional meaning of a wife looking after the household.
Grammaticalization and lexicalization seem to be continuously ongoing
processes of change. However, technically, these processes involve a large number
of successive discrete micro-step changes. Traugott and Trousdale (2010 pp. 31
and 39) label this gradualness.
Pragmaticalization is the process by which an item takes on new functions due
to contextual pressures and develops from a primarily syntactic and/or lexical
device into a pragmatic device. Of the four levels of Functional grammar, (i)
the interpersonal level (pragmatics), (ii) the representational level (semantics),
(iii) the morphosyntactic level (morphosyntax), and (iv) the phonological level
(phonology/phonetics) (Hengeveld 2008), the representational and interpersonal
levels are relevant. While an item functions at the representational level, it is still
part of the sentence grammar, yet when it starts functioning at the interpersonal
level, it becomes part of the discourse grammar. At this level, we no longer look
for clauses and sentences, but for discourse moves, the boundaries of which differ
from those of syntactic units.
An example is subordinate conjunctions functioning at the interpersonal and
discourse levels rather than the syntactic level in French. This phenomenon is
dubbed insubordination.
1 By lexicalization I mean the phenomenon that once created complex lexemes tend to become
lexical units with a specialized meaning when used frequently. Through the process of lexicalization,
the syntagmatic nature of the item in question is more or less lost (my translation).
[1] Constructionalization vs pragmaticalization
Levels que (Deulofeu 2008) parce que si (Patard

(Debaisieux 2016) 2014)
Representational universal subordination offers

circumstantial
Interpersonal discursive que noyeau-suffixe2 suggestions
Discourse que combined with an rection wishes
independent clause
At the representational level, sentence grammar still applies; at the interpersonal

level, sentence grammar no longer applies but we connect discourse moves. The
discourse level is where utterances attain their own functional profiles in the dis-
course and are not governed anymore.
In his seminal article, Evans (2007 p. 376) refuses to posit insubordination on
the reanalysis/grammaticalization spectrum. By contrast, Patard (2014) argues that
insubordination of French si-constructions is not a case of grammaticalization
but of reanalysis. Patard shows that insubordinate clauses when compared to
their subordinate counterparts are semantically modified (not bleached), syn-
tactically more independent (rather than dependent), paradigmatically add-ons
(rather than essentials), and the result of speaker economy (rather than speaker
interaction).
2.2.2 Processes of Language Contact (External)
Interferences are phenomena resulting from a language user inadvertently drawing

on two rather than one language when putting their ideas into words (cf. Chapter 1).
Interferences are unintended by the writer/speaker and thus idiolectal, yet two
people could coincidentally produce the same structure. Their being unintended
makes suggestions of strategies to avoid interferences in balanced bilinguals diffi-
cult to account for (Hamers & Blanc 2000 pp. 41–2; Torallas Tovar 2004a pp. 57–8).
One explanation for such strategies would be that an interference becomes an
established part of someone’s idiolect and recurring negative feedback on the fea-
ture in question makes the speaker aware of there being an issue with it.
2 The Approche Pronominale is a two-tier approach consisting of the microsyntax (in essence the
morphosyntax) and the macrosyntax (in essence the discourse syntax). At the macrosyntactic level,
the (verbal) noyau is the centre of the unit, a préfixe, postfixe, and a suffixe can be appended to it.
While the noyau is indispensable, the additions to it are. Examples include (i) préfixe + noyau: Comme
il était malade, il est pas venu, and (ii) noyau + suffixe: Il est venu, vu qu’il était malade (Blanche-
Benveniste 1984, 1990; Debaisieux et al. 2008 pp. 231–2).
Interferences are structures that reflect the influence of a source language on a

target language. The terms used to refer to relevant structures differ depending on
the context in which they were coined. Terms such as ‘substratum-interference’
(Matras 2009 p. 237), ‘shift-induced interference’ (Thomason 2001 p. 277) and, on
the other hand, ‘influence’ (Goglia 2005 pp. 70–3), ‘imposition (source-language
agentivity)’ (Winford 2010 pp. 170–2) and ‘transference’ (Watt 2013 p. 17) exist
(see also Hamers & Blanc 2000 p. 41). We opt for the neutral term interference in
the remainder of this study.
The source- language influence concerns the structural level; only target-
language matter appears at the surface level. In a bilingual’s mind, two languages
may interfere with each other more often than is visible in our evidence. In fact, if
a structure in language A exactly matches a structure in language B, the structure
we observe may have been selected based on the grammar of either language. For
instance, French and German employ the same preposition in [2a] and [2b]:
[2a] l’homme sur la photo.

[2b] der Mann auf dem Photo.
Hence, a French-German bilingual may have produced [2a] in a French text

based on [2a] or [2b]. Conversely, English calls for a different preposition in the
equivalent phrase:
[3a] the man in the picture.
Thus, if an English-French bilingual produced [3b] in an English text, [3b] would

be identifiable as an interference:
[3b] the man on the picture.
The above examples pertain to the distinction between positive and negative transfer
(Butler & Hakuta 2006 pp. 129–34).3 Positive transfer refers to pairs like [2a] and
[2b] in which a bilingual may have produced [2a] correctly based on [2b]. Negative
transfer refers to pairs like [2a] and [3b], in which a bilingual seems to have
produced [3b] based on [2a], but [3b] is ungrammatical. Since instances of
positive transfer are unidentifiable in our sources, we will focus on instances of
negative transfer. The concept of positive transfer may however explain an
‘unexpected impression of grammatical skill’ (Mussies 1968 pp. 74–5) in certain
cases. Mussies’s comment refers to an instance of attractio relativi in a docu-
ment that contains several ungrammatical structures.
3 Goglia (2005 pp. 70–1) dismisses this distinction based on his dismissal of interferences as
‘mistakes’.
Convergence is the process by which two varieties (e.g. dialects, sociolects,

languages) become more similar to each other (Crystal 2008 p. 113). Matras
(2010 p. 68) defines convergence in the context of linguistic patterns, that is
form-function mappings and/or structural relations between word forms and
linguistic matter (i.e. the phonological shape). He subsumes structures in which
the surface level matter comes from language A and the structural level is modelled
on language B as convergent (cf. Myers-Scotton 2002 pp. 100–5, 2006 p. 234),
including calques and metatypy.4
Muysken (2010 p. 272) summarizes convergence in language contact settings
as follows:
[4] Frequency: Relatively frequent, though not as frequent as borrowing.

Symmetry configuration: Potentially symmetrical, depending on the bilingual
community and the status of its languages.
Features involved: May lead to surface convergence, e.g. in semantic categories
and in word order.
Constraints: We do not know what constraints there are on the process; it may
be structurally quite pervasive over time. We are interested in particular in the
frequency and the symmetry of the configuration. As opposed to interferences,
convergence phenomena are frequent and may show equal input from either
source language.
Convergence of two varieties/languages may result in features being adopted

from one into the other language, yet it may also result in existing trends being
reinforced (cf. Section 4.8) (cf. Stolk 2021; Thomason 2001 p. 46).
2.2.3 Models of Language Change
Sociolinguistic approaches hold that ‘[v]ariation in language is the symbolic

expression of variation in society: it is created by society, and helps to create soci-
ety in its turn’ (Halliday & Webster 2007 p. 255). Consequently, in order to explain
language change, one has to consider three aspects: ‘the origin of linguistic vari
ations; the spread and propagation of linguistic changes; and the regularity of
linguistic change. The model which underlies this three-way division requires as a
starting point a variation in one or several words in the speech of one or two
individuals’ (Labov 1963 pp. 273–4).
4 Calques are item-by-item translations from one language into another (e.g. Wolkenkratzer/sky-
scraper) (Crystal 2008 p. 64). Metatypy is a ‘change in morphosyntactic type and grammatical organ
iza
tion (and also semantic patterns) which a language undergoes as a result of its speakers’
bilingualism in another language’ (Muysken 2010 pp. 275–6; Ross 1999 p. 7).
Sociolinguistics is a usage-based approach to language. It holds that there is an

intricate connection between the social and linguistic realities. The general idea is
that users of language are influenced by their situational surroundings including
other users of language who they come in contact with.
The seminal work is Labov (1963). Labov investigated language usage on the
island of Martha’s Vineyard near Boston, Massachusetts. He was interested in the
phonetic realization of diphthongs in English, specifically ‘the phonetic position
of the first elements of the diphthongs /ai/ as in mice and /au/ as in mouse’ (Labov
1963 p. 273).
Since sociolinguistics is an approach that focuses on the users of language and
their circumstances, aspects such as the following were relevant to Labov’s study:
i. The island has a comparatively small population but is frequented by

holidaymakers in the summer. Thus, the inhabitants interact with
speakers from different speaker communities.
ii. The islanders have a range of occupations, one of which is fishermen.
Thus, their socio-economic and educational backgrounds vary.
iii. The islanders belong to a range of age groups. Thus, they have experi-
enced and are influenced by different societal discourses—with ‘discourse’
in Foucault’s sense (cf. Gee 2014).
iv. The east of the island is more densely populated and favoured by holiday-
makers than the west. Thus, social links between inhabitants differ by
region, as does the frequency at which they come in contact with outsiders.
v. The islanders have different attitudes towards their home, Martha’s
Vineyard, and the incoming mainlanders.
Labov was interested in natural language usage. Therefore, rather than for example
asking his participants to read out sentences, he used an interview technique,
primarily to reduce the impact of demand characteristics and the observer’s
paradox, that is people manipulating how they speak in response to the experi-
mental setting. Labov found that there were features that were distinctive of
Martha’s Vineyard and that people used these more frequently when empha-
sizing their local identity, especially when contrasting themselves with others/
outsiders.
Features that were initially limited to the fishermen, the prototypical islanders,
gradually spread through social links and became a language variety specific to
the island and its inhabitants, a local dialect. We call this a bottom-up change—
that is, a feature that was originally limited to everyday registers of a small group
of workmen, spread to the entire population of the island, and in the course of
this happening the feature was reinterpreted as a social marker. The opposite phe-
nomenon is a top-down change—that is, a feature that originates in the educated
classes or literary varieties gradually spreads to the entire population and every
register; the feature in question is functionally reinterpreted in the process.5
Labov’s approach was designed for modern spoken languages, as the interview
technique shows, and for phonetic and phonological aspects of language. However,
subsequently, sociolinguistics was adapted for corpus languages and expanded
into other areas of language, including by Romaine (1982 to syntax) and by Willi
(2003 to classical Greek). Sociolinguistics has by now been firmly implemented,
further refined, and more widely explored for a range of classical languages and
text types (e.g. J. Adams 2013; Dickey 2002).
Labov’s propagation and regularization of linguistic variations is formalized by
Haugen (1966). He concludes that ‘[t]he four aspects of language development
that we have now isolated as crucial features in taking the step from “dialect” to
“language,” from vernacular to standard, are as follows: (1) selection of norm, (2)
codification of form, (3) elaboration of function, and (4) acceptance by the com-
munity’ (Haugen 1966 p. 933), as visualized below:
[5] Language and society (Haugen 1966 p. 933)
Form Function
Society Selection Acceptance
Language Codification Elaboration
At the societal level, Haugen considers the individual speaker the agent of
language change. The individual speaker produces utterances that comply with
and those that do not comply with the synchronic standard. The non-standard
structures are then either approved or rejected by the surrounding speaker
community. If structures are rejected, they remain marginal. If structures win
approval, they become integrated into the standard system (cf. Romaine 1982
p. 245). Haugen’s language level is concerned with the form-function pairing in
any linguistic feature. Form and function interact in that ‘[e]laboration of func-
tion may lead to complexity of form, and, contrariwise, unity of form may lead to
rigidity of function’ (Haugen 1966 p. 932).
What makes sociolinguistic approaches interesting for the present study is that
the individual is the centre of gravity (Matras 2009 pp. 74–9 and 310–12, 2010 p. 80;
cf. Matras & Sakel 2007 p. 849; Muysken 2010 p. 265). The individual produces
utterances due to an innate need to communicate (Matras & Sakel 2007 p. 854).
This need makes the individual produce all kinds of utterances. For the present
study, non-standard constructions that became standard are relevant insofar as
5 An example is the gradual replacement of the accusative-and-infinitive pattern by quod/quia/
quoniam clauses in Latin (J. Adams 2013 pp. 743–6). This change originated in the literary registers and
gradually spread into less formal/lower registers.
Another random document with
no related content on Scribd:
As soon as one of the head men was aware of my approach to the
king’s residence, he blew a shrill and loud blast through a long brass
trumpet, the noise of which brought all the principal male inhabitants
to the spot, who entered the hut, and seated themselves in a circle
round their sovereign and myself. The chief is a fine looking man,
apparently about fifty years of age, with a noble expression of
countenance, and a commanding air. The coozie into which I was
introduced is the largest I had ever seen in Africa, being not less, I
should think, than eighty yards in circumference. A man stood by the
side of the chief while I remained, who repeated to him what I had to
say, and the answers were returned to me by the same individual:
this singular custom is, I believe, peculiar to Beari, as I never
observed it in any other town in Africa. The chief asked the usual
questions about my king and country. The town is surrounded by a
high wall and a deep ditch, and contains about four thousand
inhabitants, some of whom had seen us before at Womba.
11th.—At six in the morning continued our journey, and arrived at
Ragada at two in the afternoon. Remained outside the walls, in order
that the horses might feed on the rich grass which grew in
abundance on the spot. The chief wished me to lodge with him in the
town; but, for the reason above stated, did not accept his invitation.
The walls of the town are about three miles in circumference.
12th.—In passing through the town this morning, at 6 A.M., waited
on the chief, who was delighted to see me, and gave me a sheep,
some fowls, and a jar of gear (a beer made from Indian corn), and
offered me a messenger, which I declined. Journeyed onwards
without halting, till we pitched our tent a little to the west of Wittesa,
at three in the afternoon. It began to rain about an hour after leaving
Ragada, and continued with great violence till after sunset. Although
the portmanteaus were defended by a thick bullock’s hide, every
thing in them, with the exception of the papers, became completely
soaked. Here I met with a party of merchants going to Koulfa.
13th.—Remained here to-day to dry our tent, clothes, &c. Boussa
Jack, the horse which was presented to my late master by the king
of Boussa, became dangerously ill. I bled him profusely, and in the
evening he was much revived. Sold an unwritten journal-book and a
few needles for 2000 cowries.
A man from the town in a state of intoxication came to my tent this
evening, with a calabash of bum, and insisted, in an insolent tone,
that I should come out and drink with him. Being busily engaged in
packing up the articles that had been put out to dry, I did not choose
to oblige him so far. When he found I had no inclination to leave the
tent, he said he was determined to come in. Wishing to intimidate the
fellow, I took a loaded pistol, and went to him, threatening to shoot
him unless he immediately left the place; but this, instead of having
the desired effect, only exasperated him the more; and flourishing a
long spear he had with him over his head, as quick as lightning he
made a desperate thrust at me. Slipping a little on one side, I caught
the weapon in my hands, within an inch of my breast, which saved
my life. I was highly incensed at this violent act, and told the people
who stood by, and were spectators of the whole affair, that if they did
not that moment take him from the tent, I would shoot him in earnest:
half a dozen of them accordingly dragged him away. Next morning
the fellow returned, and throwing himself at my feet, begged I would
not inform Sultan Magie of his conduct; for if it came to his ears he
would lose his head. I forgave him, on condition that he should never
get tipsy again.
14th.—Struck our tent at six in the morning, and passing through
the ruins of a large town called Kabojie at twelve at noon, reached
Dogo at three in the afternoon; but stopping only to pay the
accustomed duty, halted a little to the westward of it. In the evening
several respectable females of the town came to the tent, with tuah
and stewed beef, for which I gave them a few beads.
15th.—At seven in the morning proceeded on our route, and
arrived at Coulfo at two in the afternoon. I was met at the gates by
the female in whose house we lodged on our way to Kano,
accompanied by the most respectable of her sex in the town. They
expressed the most lively joy at seeing me; but when I told them that
my father was dead, they were deeply affected, and made loud
lamentations. Although the town was full of merchants, and her own
house occupied with some of them, this kind old woman turned out
her lodgers, and gave me the best apartment in it. These merchants
came from Cuttum Kora, Youri, Kano, Soccatoo, Borgou, and Yariba,
to purchase Nyffé cloth, which is the best in central Africa, iron bits
and stirrups, brass ornaments for saddles and bridles, and brass ear
and common rings: this traffic is carried on to a considerable extent.
The good old chief sent me fowls, tuah, milk, and bum: my hostess
gave me a sheep; and during my stay, this excellent old lady sent me
provisions daily.
Remained at Coulfo five days, and finding my Yariba pony, and
the bullock the king of Zegzeg had given me, unfit to bear the
fatigues of a journey to the sea-coast, exchanged the bullock and an
old Turkish jacket for a couple of asses; but could not dispose of the
pony, although I offered to sell him for a dollar. The morning before
my departure found poor little Boussa Jack dead outside my hut. It
was with the utmost reluctance my hostess would part with me; and I
left her with a promise of returning in two years. She gave me an
order to purchase looking-glasses, &c. for her on my arrival in
England, which I was to bring with me; all which I promised to do.
Gave her a dollar to make a ring, and half a yard of scarlet damask. I
took to the old chief a scarlet cap and one of my late master’s
turbans, and begged he would not compel me to visit Sultan Magie
at Sanson, as I had nothing to give him. He replied, if I did not wish
to visit the sultan, to leave as soon as possible, or he might send
messengers for me. Sold 5000 cowries worth of needles to assist me
on my journey.
20th.—Having paid my respects to the king last evening, left
Coulfo, in high spirits and good health, at six in the morning, and
halted outside Makonja at five in the afternoon. Here I found a party
of merchants going to Gonja, near Ashantee, for gora nuts. Unable
to procure any kind of provisions in the town, I took my gun and shot
a heron; not being aware that this, being a fetish bird, is allowed to
be eaten only by the chiefs. On being made acquainted with the
circumstance, one of the principal inhabitants came with a loaded
gun, and threatened to shoot me, making use of the most abusive
language. I dared him to fire, and said I should complain to Sultan
Magie of his conduct.
21st.—Early this morning, the person who had threatened to
shoot me yesterday came to beg my pardon, excusing his conduct
on the plea of being intoxicated with drinking too much bum. At six in
the morning continued our route, and at twelve at noon halted at
Dalho, on the banks of the river Quontakora. At 3 P.M. every thing
was safely landed on the opposite side; and journeying till sunset,
reached the foot of a hill five miles to the east of the Niger, and
rested there for the night. The country traversed to-day was one
continued swamp, the mud and water frequently reaching up to the
asses’ side. Rained continually throughout the day.
22d.—At nine in the morning left the hill, and at eleven at noon
came to the river Niger, which was full one hundred yards wider than
we found it on our way to Soccatoo. After waiting an hour, a canoe
came to fetch us; but we did not get over till two in the afternoon, the
horses, &c. being obliged, by reason of the rapidity of the current, to
be ferried across in the canoe. My poor slave Jowdie, never having
in all his life seen so large a body of water before, became greatly
agitated, and was afraid to cross. Getting him at length, after much
persuasion, into the canoe, he reached the other side, but fainted the
moment he put his foot on terra firma. We landed at Inguazhilligee,
where the principal inhabitants pressed us to remain; but having no
presents to offer, declined stopping. Expecting to reach a village in
the evening, made all possible haste to get out of the town; but at six
o’clock, P.M. the asses being quite spent, were obliged to fix our tent
on a marshy spot in the midst of a bush. The roads were horribly
bad, through swampy ground, and could only travel two miles an
hour. This evening, as well as the two former ones, went to rest
without food.
23rd.—At six in the morning, proceeding on similar roads, and
having heavy rain, reached a small village at one, and were obliged
to remain there for the day to give the asses a little rest. The poor
animals had stuck in bogs seven times in the morning, each of which
times the loads were taken from their backs, and they themselves
dragged with much difficulty and loss of time out of their unpleasant
situation. None of the poor inhabitants of the village understanding
the Houssa language, I entered a hut, the door of which was open,
and giving the inmates a few beads, made signs with my hands that
we stood in need of provisions and rest. One small lean fowl,
however, was all that could be procured from them; but having the
good fortune to shoot a guinea-fowl, made a tolerable meal, and
slept soundly.
24th.—At seven in the morning left the village, and journeying till 3
P.M., halted at the pleasant town of Wowow, one of the handsomest,
if not the handsomest, in the interior of Africa. It had taken us two
days and a half on the road from Inguazhilligee, while in the dry
season we did it with ease in one.
25th.—Took the old king a present, consisting of four yards of
scarlet and four yards of blue damask, four yards of scarlet silk, a
cap of the same colour, and a white turban. Though overjoyed to see
me at first, he was much distressed on mentioning the death of
Captain Clapperton at Soccatoo. He wondered at my being alive,
after visiting the Fellata country, the inhabitants of which he
described as the most barbarous in the world; said my preservation
was miraculous, and that I was certainly one of God’s favourites. I
was obliged to comply with his request, which was tantamount to a
command, to remain with him a few days to clean seven muskets
and three pistols, which he afterwards told me belonged to the white
men who were drowned at Boussa. They have the Tower mark on
them.
26th.—Solicited and obtained permission to visit Boussa, to pay
my respects to its king.
27th.—At five in the morning, accompanied by Pascoe,
proceeded on the road to Boussa; but, after journeying till six in the
evening, had only got half way. The country is full of bogs and
swamps, which in some places took the horses up to their saddles in
mud and wet; and Pascoe being the worst rider I ever saw, was
soused in them more than a hundred times. The remaining part of
the road appearing still more difficult and dangerous, found it
impossible to proceed any further.
28th.—Returned to Wowow.
29th.—The king sent me a goat, cut up into small pieces, and a
large bowl of tuah, to make a sadacco for my poor master; a
ceremony common in many places in the interior, on the decease of
any person of consequence. The pieces of goat and the bowl of tuah
are sent to the mallam or priest, who repeats a short prayer over
them: it is however necessary, before this can be performed, to place
a gold or silver coin, or at least an article made of either of those
metals, on the top of the bowl. Having no coin I could part with, I put
a silver pencil-case in its stead, which was never returned me. After
the prayer has been pronounced over the tuah and goat’s flesh, they
are sent back to the person to whom they belong; and any one who
is disposed comes and eats; each person, before tasting, repeating
this pious ejaculation, “God send him safe to heaven!” The goat’s
flesh and tuah sent to me were soon consumed.
30th.—Having finished cleaning the muskets and pistols, asked
the chief permission to leave Wowow. The old man, smiling, told me
not half my business was done; he wanted six charms, which I alone
could write. These charms were to be worn on his person, and to
possess the following virtues: 1st charm. If his enemies thought of
making war on him, it would cause them to forget to put it in practice.
2d. If they should be on their way to his city, for the purpose of
warring, it would turn them back. 3d. If they should discharge their
arrows at his people, when close to the city walls, it would cause
them to rebound in their own faces, and wound them. 4th. It was the
province of this charm to prevent his guns from bursting. 5th, was to
preserve the person who might hold the gun from receiving any
injury, should it unfortunately explode. The 6th and last charm was to
make him the happiest and most successful of men.
31st.—Carried the charms to the king, on which I had written
scraps of old English ballads, which made him in the best humour in
the world.
September 1st.—The king still wished to detain me longer, and
insisted on my selling or giving him my gun and pistols, the only
arms I had left. I endeavoured to soften him by every means in my
power; but finding him firm in his determination of having the gun,
and at least one of the pistols, and knowing resistance would be
ridiculous, I sent them to him, leaving it entirely to his generosity to
give what he thought proper. The liberal-minded chief shortly
afterwards returned 4,000 cowries (worth little more than a dollar), as
a sufficient remuneration for the gun and pistol, but made me a
present the next day of a beautiful little mare.
3d.—Early in the morning the old chief desired to see me before I
left, and obtained from me a promise to return to him after having
visited my own country. He showed me various patterns of silk for a
tobe I was to bring from England for him; and said, as I was going
out of the apartment, “Your countrymen may come here, and build a
town, and trade up and down the Niger: we know now that they are
good men, but we did not think so when the white men who were
drowned at Boussa were in the country.” He kept me with him till
nine o’clock in the morning, when, on going to my hut, I found a party
of merchants waiting for me, whom the kind old chief had detained
on purpose that they might accompany me to Khiama (which had
been concealed from me entirely), the roads to that city being
infested with bands of robbers. Crossed the river Auli at twelve at
noon. The current being very rapid, we had extreme difficulty in
getting over; but no accident occurred, and we fixed our tent on the
south bank of the river. In the evening the mallam, or priest of the
merchants, came to my tent, and gave me the following account of
Mungo Park and his unfortunate companions:
“You are not, Christian, the first white man I have seen. I knew
three of your countrymen very well. They arrived at Youri at the fast
of the Rhamadan (April). I went with two of them three times to the
sultan. The person that appeared to be the head of the party made
the sultan a valuable present on one of his visits, which consisted of
a handsome gun, a cutlass, a large piece of scarlet cloth, a great
quantity of beads, several knives, and a looking-glass. He was a
very tall and powerful man, with long arms and large hands, on
which he wore leather gloves reaching above the elbows. Wore a
white straw hat, long coat, full white trousers, and red leather boots.
Had black hair and eyes, with a bushy beard and mustachios of the
same colour. The sultan of Youri advised your countrymen to
proceed the remainder of the way on land, as the passage by water
was rendered dangerous by numerous sunken rocks in the Niger,
and a cruel race of people inhabiting the towns on its banks. They
refused, however, to accede to this, observing that they were bound
to proceed down the Niger to the Salt Water.” The old mallam further
observed that, “as soon as the sultan of Youri heard of their death,
he was much affected; but it was out of his power to punish the
people who had driven them into the water. A pestilence reaching
Boussa at the time, swept off the king and most of the inhabitants,
particularly those who were concerned in the transaction. The
remainder, fancying it was a judgment of the white man’s God,
placed every thing belonging to the Christians in a hut, and set it on
fire.” It is not a little remarkable that it is now a common saying, all
through the interior of Africa, “Do not hurt a Christian, for if you do,
you will die like the people of Boussa.” The old man left me shortly
afterwards; and I thanked him for his information thus voluntarily
given.
4th.—At nine in the morning left the bank of the river, and at two in
the afternoon fixed our tent in a bush. Proceeded this day by
ourselves, the merchants not being ready to start in time.
Notwithstanding the account the king of Wowow had given me of the
dangers of the road, we were not molested in our journey.
5th.—At six in the morning were again in motion, and halted at
four in the afternoon in a small town called Gorkie. This day one of
the asses was taken suddenly ill on the road, and it was with the
utmost difficulty we could drag him to the halting place. The chief of
the town, at the instigation of the king of Wowow, supplied us
abundantly with fowls, rice, yams, &c. Gave him a pair of scissors
and a few needles.
6th.—Finding the ass that was taken ill yesterday unable to
proceed any farther, was obliged to leave him behind. On seeing his
companion going away, the poor animal brayed most piteously, and
made many ineffectual attempts to rise. At six in the morning left the
town, and after sunset came to a number of grass huts, similar to
those mentioned before; in the largest of which we rested for the
night, and did not fix the tent. On crossing a river, apparently very
narrow, to-day, my horse unexpectedly sunk several feet in mud and
sand, and his legs becoming entangled in the roots of some trees
lying at the bottom, fell on his side, and plunging violently, threw me
off his back; but not being able to get one of my feet from the stirrup,
lay several seconds under water, in a most perilous and distressed
state. Freeing myself at length from the stirrup, I got out much
exhausted, and succeeded in extricating the horse also from his no
less unpleasant situation. Some merchants told me afterwards at
Khiama, that I might think myself highly fortunate in escaping so
easily, the river being filled with large crocodiles.
7th.—Started at six in the morning, and arrived at Yaro after
sunset. The chief made us the usual presents of provisions, &c. A
dreadful hurricane came on about nine o’clock in the evening, which
lasted three hours, when it became awfully still and calm. The storm
was unusually violent, and tore up large trees by the roots; the peals
of thunder were fearfully loud, and shook the ground on which our
tent was fixed; while large masses of fire, falling from the heavens,
added not a little to this scene of devastation and terror.
8th.—Heavy rain all day; unable to proceed.
9th.—At half-past six in the morning continued our journey, and at
noon entered the city of Khiama. Went immediately to the king’s
residence, who, as soon as he saw me, asked how I dared to come
into his town without having previously sent him a messenger to
inform him of my approach. I answered I had sent one of his own
men, three or four days before, to acquaint him of it. “That is of no
consequence,” he continued; “you should have sent another this
morning. Get on horseback directly, and return an hour and a half’s
journey the way you came: on arriving there, send me a messenger,
and I will order a sufficient escort to conduct you into the city, in a
manner deserving your rank and respectability.” I was in the act of
obeying this odd mandate, when he bawled after me: “I forgive you
this time, Christian, but never be so remiss again.” The chief is an
eccentric but friendly kind of man, and regretted very much my
father’s death, which had been told him some days before by a
merchant. He had heard the Fellatas had behaved very roughly to
us, and had robbed us; but asked what business we had at
Soccatoo? I told him we were at Kano, on our road to Bornou, when
Sultan Bello sent for us, and we were obliged in consequence to visit
him. He offered to send me safely to Bornou, and said he was
tributary to the sheik. I replied, that Bello had taken the presents
intended for him, and that I had none left worthy so great a prince,
therefore could not accept his kind offer. My present to the chief
consisted of a silk sword-sash, three yards of scarlet and blue
damask, the same quantity of blue silk, a red cap, two pairs of
scissors, and a hundred needles. I likewise gave him my old tent,
which was full of holes, and quite useless.
Remained at Khiama five days, during which time the queen’s
treatment was kind and generous, supplying us every day with
excellent provisions in great abundance. The day before my
departure the king gave me a strong pony; and observed that, “if my
king wished, at any future period, to send any one to Bornou, he
would conduct him there by a safe route, without the necessity of
going through the Fellata country.” The king of Khiama is, without
exception, the finest and handsomest man we had seen in Africa (far
superior to Bello); and, with the exception of the king of Yariba, was
most respectably dressed.
14th.—At six in the morning left Khiama, and halted in Subia at
one at noon. The country traversed is full of bogs, which caused the
horses to fall repeatedly with their riders. The chief of this town gave
me a goat and some yams, by command of the king of Khiama’s
messenger.
15th.—Started at six in the morning, and travelled the whole of the
day, till ten at night, when we rested in a bush.
16th.—Crossing a creek at 6 A.M. arrived at Mossa, a town
situated on the banks of a river of the same name, which divides
Yariba from Borgoo. The river was overflown, and the current strong
and rapid; in consequence of which the people of the town were
afraid to ferry us over. No food of any kind this day.
17th.—Asked the king of Khiama’s messenger why he was so
much afraid of crossing the water, observing that I had myself swam
across many larger and more rapid rivers; and among others
mentioned the Niger. The man, in great trepidation, begged me, as I
valued my life, not to mention the names of rivers in the hearing of
the Mossa, who was a female river, and had many rivals in the
affections of the Niger, who was her husband. She had a capricious,
jealous, and cruel disposition; and if I ventured to place myself in her
power, she would certainly swallow me up, as I had spoken
slightingly of her. She was continually quarrelling with her husband,
thinking he was too familiar with other rivers; and where they met
they made the “devil’s own noise” with their disputes. I roared with
laughter when the man had done speaking, at the loves of the Niger,
which made him very angry, and I had much to do to pacify him.
Being unable to procure provisions of any kind in the village, went to
the chief, and wished to know whether he intended to starve us. The
old scoundrel had a garden in which he grew a quantity of yams; but
he refused to sell me any, asserting that he had none for himself. I
then asked leave for Pascoe to cut grass in his garden for the
horses, suspecting his yams might be hid in some part of it: this he
sulkily granted. In the evening Pascoe returned with a bundle of
grass, concealed in which were several yams he had had the good
fortune to discover. If this had not been the case, I really believe we
should all have died of hunger.
18th.—Went again to the chief in the morning, and demanded
something to eat. He solemnly declared he was himself starving, and
could get nothing to eat. I then pressed him to send his canoe to the
opposite side for provisions: after much intreaty he consented to do
so, but first made a charm that it might not be injured in its perilous
voyage. For this purpose he killed a fetish fowl, sprinkled its blood in
the river, placed some of the entrails in the bow of the canoe, and in
the stern put a broken egg: he then muttered several expressions I
did not understand; and a man was sent over with the boat, which
arrived there in safety; but in returning loaded with poultry and yams,
it ran foul of a tree which unluckily lay in the middle of the stream,
and sunk immediately, in consequence of which the whole cargo was
lost. The chief informed me of the accident, and said that the river
could not be crossed in safety for three or four days. Pascoe
returned in the evening with more of the chief’s yams and grass.
19th, 20th, and 21st.—Nothing to eat on these days but the yams
Pascoe stole from the king’s garden.
22d.—Attempted to cross the river at noon; but did not succeed in
getting every thing on the opposite side till five in the evening. The
horses and asses were borne a quarter of a mile down the stream,
and had nearly perished. Halted at Wantatah, the first town in Yariba,
about two hours after.
23d.—At seven in the morning pursued our journey, and at eleven
entered Hogie, where we remained for the day. The chief gave me a
pig, yams, corn, &c. A red cap and fifty needles fully satisfied him.
24th.—Rested at Hogie to-day.
25th.—Left at six in the morning, and arrived at Katunga, the
capital of Yariba, at seven in the evening. The low grounds were
rendered almost impassable by reason of the rain, which fell in
torrents. On our arrival I was put into the same house we occupied
on our journey into the interior.
26th.—The king would not let me wait on him, fearing it might wet
my feet; and accordingly he visited me with five hundred of his wives
(out of two thousand), and the principal inhabitants of the city. The
wives welcomed my return by singing a simple and plaintive air, with
much pathos and feeling: their voices were sweet and musical; and
the whole had a novel and pleasing effect. Nothing could be heard
but their strains, to which every one listened with the profoundest
attention until the conclusion of the performance. The king
expressed his sorrow for my master’s death; and questioned me
very minutely on the motives that induced us to go into the interior.
On telling him it was to see if there was any thing worth trading for in
the country, he appeared satisfied. He was richly dressed in a scarlet
damask tobe, and a pair of trowsers made of country cloth, scarlet
ground with a blue stripe; the former ornamented with coral beads;
his legs, as far as the knee, were stained red with hennah; and on
his feet he wore red leather sandals. A cap made of blue damask,
thickly studded with coral beads, was on his head; and silver rings
hung round his neck, arms, and legs. I offered him the horse I had
purchased at Kano, a fine animal, that had carried me the whole of
the way from that city: and regretted my inability to make him a more
valuable present; but promised, if he permitted two messengers to
accompany me to the sea-coast, I would send him something else.
In the evening I received a goat and a great quantity of yams from
the king.
27th.—The king having desired me to call on some of his head
men, I waited on the master of the horse, and two others. The former
gave me a goat and a bottle of honey. I remarked I was very poor,
and could make him no return.
30th.—I informed the king I was short of money, on which he
generously sent me a duck and four thousand cowries (little more
than a dollar). The eunuch, the king’s head man, begged of me my
remaining pistol, two dollars, and a scarlet cap, which I was
necessitated to give him. He also wanted my ass to make a fetish,
but this I refused to let him have. This evening I sent for the ass from
a neighbouring pasture, when I found the poor animal had been shot
in the side with two poisoned arrows. I have no doubt but this cruel
action was performed at the instigation of the disappointed eunuch.
The beast became a complete skeleton, and after languishing for six
days in great agony, I desired Pascoe and Jowdie to take him to a
short distance, and cut his throat. When the king heard of the
circumstance, he immediately ordered the carcass to be cut into
quarters, and conveyed to his house. After which he ordered the
meat to be dressed, and having assembled his wives and head men,
they regaled themselves on it with peculiar satisfaction. Wishing to
pay for so delicious a treat, the king sent me a goat and a thousand
cowries for the dead ass.
The people of Yariba are not very delicate in the choice of their
food; they eat frogs, monkeys, dogs, cats, rats, mice, and various
other kinds of vermin. A fat dog will always fetch a better price than a
goat. Locusts and black ants, just as they are able to take wing, are
a great luxury. Caterpillars also are held in very high estimation. The
caterpillars are stewed, and eat with yams and tuah. Ants and
locusts are fried in butter, and are said to be delicious. I could never
make up my mind to taste any of these rich insects; Pascoe,
however, is particularly fond of them, and calls them land-shrimps.
It is a custom in Katunga, when the king dies, for his eldest son,
first wife, and all the head men of the kingdom, to drink poison over
his grave, and are afterwards buried with him. None of the king’s
sons ever come to the throne. After the king’s death, his successor is
chosen from among the wisest persons of the country; an elderly
man is generally preferred.
Remained at Katunga till the 21st of October, when the king gave
me 4,000 cowries, and some trona to sell on the road. He ordered
his head messengers to accompany me, with a desire they should
command the chiefs of every town through which we were to pass,
to contribute, according to their means, to our support.
On the 22d of October I left Katunga, and after a rapid and hasty
journey, during which nothing of consequence occurred, arrived at
Engwa on the 9th of November following. I found the railing which
had been placed round the grave of poor captain Pearce washed
away by the rains, and the piece of wood on which I had myself cut
out his name, age, &c. and the day on which he died, either taken off
by the natives, or also washed away by the rains. The only means by
which I recognized the spot was the appearance of the earth, on
which there was no vegetation, and its having sunk five or six inches.
I went to the chief, and giving him half a yard of blue damask, and a
pair of scissors, entreated him to build a house over the spot, and
have it thatched, promising to send him a present from England in a
short time, if he did it. He promised faithfully to erect a house as
soon as the rains were over.
Nov. 12th.—Arrived at Jennah this afternoon, and found the grave
of Dr. Morrison to be in a perfect condition. The king had had it kept
in excellent order, and was to have been rewarded by Mr. Houston,
who himself died shortly afterwards at Accra. In this city the horses
the kings of Wowwow and Khiama had given me, unfortunately died.
Remained two days at Jennah, when we proceeded, and on the
21st entered Badagry; all the natives on the road from Kano having
behaved remarkably well to us. The country traversed was swampy,
and full of water the greatest part of the way, which rendered
travelling tedious and unpleasant.
The king was glad to see me, gave me his own house,
constructed of bamboo, whilst he himself, much against my
inclination, resided in a small miserable mud hut. Like every other
prince on the road, he was grieved to learn the decease of my
master. I gave him my little horse, which I had brought the whole of
the way from Soccatoo, and all the articles for presents that were
left, consisting of two yards of light blue damask, and the same
quantity of scarlet ditto, two yards of scarlet and blue silk, and two
dozen pair of stockings.
On the 28th I waited on captain Morrison, a Portuguese slave-
merchant, and obtained from him goods amounting to ninety-four
dollars; and a barrel of gunpowder, which cost a doubloon, I paid for
myself.
Three of the Portuguese slave-merchants residing at Badagry,
went to the king one day, and told him and his principal men that I
was a spy sent by the English government, and if suffered to leave,
would soon return with an army and conquer their country. This the
credulous people believed, and I was treated with coldness and
distrust by the king and his subjects, who seldom came to see me.
All the chief men at length assembled at the fetish hut, and having
come to a resolution that I was to drink a fetish, sent for me to
appear before them. On my way five or six hundred people gathered
round me, and I could proceed with difficulty. A great number of them
were armed with hatchets, bows and arrows, and spears; and waited
outside the hut till I came out. On entering one of the men presenting
me with a bowl, in which was about a quart of a liquid much
resembling water, commanded me to drink it, saying, “If you come to
do bad, it will kill you; but if not, it cannot hurt you.” There being no
resource, I immediately, and without hesitation, swallowed the
contents of the bowl, and walked hastily out of the hut, through the
armed men, to my own lodgings, took powerful medicine and plenty
of warm water, which instantly ejected the whole from my stomach,
and I felt no ill effects from the fetish. It had a bitter and disagreeable
taste, and I was told almost always proved fatal.
When the king and chief men found, after five days, that the fetish
had not hurt me, they became extremely kind, and sent me presents
of provisions, &c. daily, and frequently said I was protected by God,
and that it was out of the power of man to do me an injury.
I remained at Badagry two months, but was advised by the king
never to go out unarmed, as the Portuguese took no pains to
conceal their inveterate dislike to me; and would no doubt
assassinate me the first opportunity that might present itself. Several
persons in canoes passed repeatedly from Badagry to Cape Coast,
but although I offered to reward them, I could never get them to take
a note for me to the latter place, so active had been the exertions of
the Portuguese to debar me from every means of communicating
with my countrymen.
There were five factories at Badagry, in which were upwards of
one thousand slaves of both sexes, chained by the neck to each
other, waiting for vessels to take them away.
Captain Morris, of the brig Maria, of London, hearing of my being
at Badagry, kindly came from Whydah to fetch me, and on the 20th
of January I went on board, and arrived at Cape Coast on the 31st of
the same month. Here I gave my faithful slaves, Aboudah, Jowdie,
and Pascoe’s wife their freedom, who testified their sorrow on my
departure by heaping sand on their heads, and other marks of grief
peculiar to the African race. Colonel Lumley generously promised to
give them pieces of ground and a small sum of money, and I have no
doubt they will do well.
Sailed from Cape Coast in the Esk sloop of war February 3d. and
arrived in England the 30th of April following.
THE END.
APPENDIX.
A LIST, OR SUMMARY ACCOUNT, OF THE LATE CAPTAIN

CLAPPERTON’S ARABIC PAPERS, TRANSLATED BY MR. A. V.
SALAMÉ.
1. A geographical description of the course of the river

Cówara, the road from Soccatoo to Maséna, from Timbuctoo to
Soccatoo, and the countries thereabouts, with their inhabitants,
produce, &c. &c. as delineated in the chart annexed.
2. An account of an expedition of discovery, made some years
ago, by forty Christians, who had built a ship or vessel in the interior
of Africa, and proceeded down the river Cówara.
3. A geographical account of the country, rivers, lakes, &c. from
Bornou to Egypt, Nubia, Sennar, the Nile, its sources, &c. &c.
4. A traditional account of the people and country of Malí, &c. &c.
5. A traditional account of the origin of the Felan tribe, whom we
have hitherto erroneously called “Fellatahs.”
6. A traditional account of the people and country of Bargho, &c.
&c.
7. An Itinerary, or the road from Sérá to Nubia, Sennar, and Egypt.
8. A traditional account of the people and country of Nefí, &c. &c.
9. A traditional account of the people and country of Noofy, &c.
&c.
No. I.
The following is a translation of the Arabic which was written on
the original chart (as Clapperton calls it), and of which the annexed
is a reduced copy, professing to describe the course of the Cówara
river, the road from Soccatoo to Maséna, and from Timbuctoo, with
the names and geographical description of the towns and countries
adjoining.
A. “Representation of the city of Soccatoo, metropolis of the

Prince of the Believers (Sultan Bello).”
B. “The island and town of Oodel or Goodel, together with the
branch of the river which surrounds them.”
C. “The town of Boory, two days’ journey from Bagrá-foogal.”
D. “The town of Bagrá-foogal, one day’s journey from Ghagró.”
E. “The town of Ghagró, three days’ journey from Toondebí.”
F. “The town of Toondebí, one day’s journey from Sharif.”
G. “The town of Sharif, three days’ journey from Kasbi, or Kasb.”
H. “The town of Kasbi, two days’ journey from Seegho, or
Sheeghó.”
I. “The town of Seegho, or Sheeghó, one day’s journey from
Kabará.”
J. “The town of Kabará, half a day’s journey from Timbuctoo.”
K. “The city of Timbuctoo, ten days’ journey from Jéri.”
L. “Representation of the city of Timbuctoo.”
M. “The island and town of Jéri, together with the branch of the
river that surrounds them. This island is in the middle of the territory
of Maséna, or Masera, and between it and Sego is seven days’
journey.”
N. “This branch of the river is called, in the language of the Felan
people, Bálió, and in Arabic the Black Sea or River. It extends from
Jéri to Foota-Jaló.”
O. “This branch is called by the Felans Ranioo, and in Arabic the
White Sea or River. It extends from Maséna to Sego, Foota-toroo,
and to Darhoz (perhaps San Salvador), one of the towns of the
French Christians.”
P. “The road from Soccatoo to Maséna, which is frequented in
these days, across the river and the island of Oodel, with the names
and description of the towns and countries that are on it, between
the river and Maséna.
“On the west bank of the river the country of Biténkoobi lies. Its
inhabitants are of the tribe of Felan; it abounds with mountains,
rocks, plains, elephants, and buffaloes; and has along the bank of
the river many white hills. Some of the people drink their water from
the river, and others have shallow wells.
“Next to this, at one day’s journey through an inhabited country,
the territory of Maázo-moudi lies, whose inhabitants are of the tribe
of Tooroodi. It is a low mountainous country, and abounds with
thorny trees and wells. Its prince is a tall and extremely strong man,
and renowned for his courage and wars.”
Q. “Is the country of Yaghrá; between it and Maázo-moudi is
about three days’ journey, through desert and stony plains, with a
few trees and some mountains. In the midst of the plains, however,
there is a well-known river, called Sirba. Yaghra is now possessed by
a prince of the Felans, named Ibrahim (Abraham) Boonti. It contains
woods, small mountains, and a well-known deep river, called Yalí,
from which the people get their water.”
R. “Is the country of Lebtakó; between it and Yaghrá is two days’
journey, through woody plains and low ground. Its inhabitants are
Felans, and its prince is named Saléh. They are a great warlike
people, possessing fine swift horses and many cattle. They subsist
upon the corn called dokhun (millet); and their country is hilly and
sandy, and has a large lake, called Dúry.”
S. “Is the country of Jelghooji; between it and Lebtakó is two days’
journey, one of which is through villages, and the other through
desert plains, in the midst of which there is a large lake, called
Bookma. The king of this country is named Hamarkoli, and the
people are Felans, and well known as great warriors. They possess
abundance of swift horses, oxen, and other cattle. The country is
mountainous, woody, and has a well-known lake, called Jeboo,
besides a great many wells.”
T. “Is the territory of Hajrí, one day’s journey from Jelghooji. It is
extensive, but very rocky, stony, and mountainous; has a great many
sandy hills and a few vales; water in it is very scarce, and it has but
few wells, so much so, that its inhabitants are obliged to store their
water, during the rainy season, within the trunks of trees. The
inhabitants of the vales are the Felans, who originally conquered
these countries; but the mountains are inhabited by a people called
Benoo-Hami (the children of Ham), of the tribe of Sokai. They are
great warriors, subsist upon the dokhun, and have abundance of
swift horses and oxen. In the middle of this country there is a very
great and lofty mountain, equal to which is not known in those parts,
and here is its representation. [See ◬ in the chart.] Upon it there is a
town called Oonbori, whose king is named Noohoo-ghaloo-farma, of
the tribe of Sokai, and is renowned for his generosity and
munificence.
“All these countries, except Oonbori, are subject to our Lord the
Prince of the Believers, Mohammed Bello, whom may God cause to
be ever victorious, for the glory of the faithful, and the annihilation of
the infidels.”
U. “Is the territory of Maséna, seven days’ journey from Hajrí: it is
very extensive, fertile, and abounds with rivers and lakes. Its villages
join both the old and modern ones of the Felans. Its inhabitants are
powerful warriors, since ancient times possessing abundance of
oxen and sheep, and are blessed with every comfort of food and
living; most of them subsist upon rice, butter, fish, and meat. In the
middle of the country there are two lofty mountains, called Soroba
and Goran, and its cattle feed, throughout the four seasons of the
year, upon grass; and as the lands are almost continually covered
with water, the shepherds and herdsmen gather the grass, heap it up
in large stacks, and, while the cattle are grazing, live upon the tops

Coptic Interference in The Syntax of Greek Letters From Egypt Victoria Beatrix Maria Fendel Full Chapter

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Coptic Interference in The Syntax of Greek Letters From Egypt Victoria Beatrix Maria Fendel Full Chapter

Uploaded by

Copyright:

Available Formats

Coptic Interference in the Syntax of

Greek Letters from Egypt Victoria

Great Clarendon Street, Oxford, OX2 6DP,

To my parents Andreas and Brigitte

I am deeply grateful to my DPhil supervisor, Andreas Willi, who guided me all

Becher-­Häusermann (Basel), Roxanne Taylor (Oxford/Manchester), Alexander

Victoria Beatrix Maria Fendel

5. Verb Phrases: The Syntax of Arguments 237

10. Summary and Conclusion 445

Appendix: Corpus of Texts 485

A Achmimic dialect of Coptic

xii List of Abbreviations

L Lyko-­Diospolitan dialect of Coptic

DDbDP Duke Database of Documentary Papyri (http://papyri.info)

1.1 Bilingual Interference

Bilingual interference refers to a language user inadvertently drawing on two

[1a] the man on the photograph.

[1b] l’homme sur la photo.

4 Coptic Interference in the Syntax of Greek Letters

in both languages often discovered together. Linguistically, there are bilingual

[2] PKC 19, ll. 13–14 (Kellis, 4th c. ad)

1.2 Egyptian Interference in Greek

period. Additionally, Gignac (2013) hints at several structures where he suspects

1.2.2 Multiple Influences

In her contribution to A Companion to the Ancient Greek Language, Torallas

6 Coptic Interference in the Syntax of Greek Letters

Conversely, in his description of the early Byzantine period in The Oxford

[4] P. Ross. Georg 4 app. pp. 99–105 (Aphrodites Kome, Antaiopolites,

5 CG ἦ(ν) ‘he was’ (Jernstedt 1927).

The following phonetic spellings appear in the letter:

• iotacism (οι/η/ει/υ), e.g. l. 4 μνίμιον for μνημείων;

The letter exhibits various features of post-­classical morphology and syntax.

8 Coptic Interference in the Syntax of Greek Letters

[5a] l. 1 πολὰ κόπον ἔπαθα

[5b] ll. 3–4 ἀπέθανα ὁς ἕναν ἐκ το῀ ν μνιμίον

Jernstedt (1927 p. 103) considers ὁς (ὡς) equivalent to MG σάν (ὡσάν), a prep­os­ition

[5c] l. 5 ἐξήλισα εἰς τὰς χύρας αὐτο῀ ν

1.2.3 Morphology, Syntax, Phraseology

9 This is Torallas Tovar’s (2007) interpretation of the terminology.

10 Coptic Interference in the Syntax of Greek Letters

(1) When the numeral ‘two’ functions as a quantifier in combination with a

[8] ὁ ἀνὴρ ἥν ὁρᾷς

10 The writer repeats the construction in l. 24.

[9] οἱ παῖδες τῶν πατρός

[10] Sophocles, Antigone 1270–2

[11] (Till 1936 p. 11, ll. 9–11)

12 Coptic Interference in the Syntax of Greek Letters

ē nim m-­monakhos pe nta-­f-­bōk šaro-­f etbe p-­oudjai n-­tef-­psukhē

All these structures are too widespread in post-­classical Greek to be classified as

[12] {ὁ Παῦλος}sender {τῇ Μαρίᾳ}addressee {χαίρειν}greeting

[13a] ὅρος ⲧⲟⲟⲩ toou ‘mountain’ and ‘monastery’.

1.3 Guiding Principles and Outline

14 Coptic Interference in the Syntax of Greek Letters

Chapter 2 comprises a thorough introduction to bilingualism and language

1.4.1 Clusters and Interdependencies

16 Coptic Interference in the Syntax of Greek Letters

[14a] AP 1, ll. 19 καὶ τὸν μονάριν Ἡρακλίδην̣ δύ[σα]ντες καὶ ὑβρίσαντες

In [14a], orthographic variability impacts on which verbal lemma we catalogue.

[14b] AJ 7, ll. 7 ἵνα ἀπολύομαι

In [14b] and [14c], orthographic variability affects syntactic correctness. If we

[14d] AN 13, ll. 11]πραὺ̣ ἐπιξαμε[ν

In [14d], orthographic variability is relevant to the voice of the verb. If we read

1.4.2 Structure, Context, Formation

Becher-Häusermann (Basel), Roxanne Taylor (Oxford/Manchester), Alexander

Appendix: Corpus of Texts 485

L Lyko-Diospolitan dialect of Coptic

The letter exhibits various features of post-classical morphology and syntax.

Jernstedt (1927 p. 103) considers ὁς (ὡς) equivalent to MG σάν (ὡσάν), a preposition

ē nim m-monakhos pe nta-f-bōk šaro-f etbe p-oudjai n-tef-psukhē

All these structures are too widespread in post-classical Greek to be classified as

Lexical-Functional Grammar (henceforth LFG) is built around three

15 A recent up-to-date reassessment of LFG is Dalrymple et al. (2019).

Drawing on three clear-cut diagnostic criteria homogenises the assignment of

Structures that have resulted from interference or a proficient individual’s idiolect

Hence, a French-German bilingual may have produced [2a] in a French text

Thus, if an English-French bilingual produced [3b] in an English text, [3b] would

Sociolinguistics is a usage-based approach to language. It holds that there is an