“Genitive” possessive constructions in Égyptien de tradition: Compound construction vs. of-construction Daniel A. Werning (EXC “Topoi,” Humboldt-Universität zu Berlin) Keywords: Egyptian, Égyptien de tradition, attributive possession, noun compound, of-construction, emulation of ancient language, passive possession, animacy, alienability, genitive meaning, semantic map, phrase classifiers. Abstract This article mainly explores the usage of two attributive possessive constructions that are traditionally labeled the “Direct Genitive” and “Indirect Genitive” (in both cases, the notion “genitive” refers not to a morphological case but to an attributive possessive function). The former is a productive type of noun compound construction, the later is comparable to the English of-construction. The texts investigated are written in Égyptien de tradition, i.e., in an artificial emulation of “ancient” language in the later 2 nd and 1 st millennia BCE. Among other things, the article explores constructional as well as semantic influences on the choice of one or the other construction in the Book of Caverns (13 th century BCE): (i) influences of constructional complexity of the two related nouns/noun phrases, (ii) the animacy of possessors, and (iii) the alienability of the possessive relation. In the course of this, the statistical impact of cases of “agentive possesseds” like, e.g., ‘ruler of the netherworld’ and “possessing possesseds” like, e.g., ‘possessor of a bier,’ is discussed, i.e., the impact of cases in which the possessed “controls” or possesses the possessor. Finally (iv), the attested “genitive” meanings are mapped onto a semantic map of possessive relations. The quantitative observations made are interpreted as the result of the application of a simple translation rule by which the ancient authors transformed genitive constructions of their contemporary language varieties into genitive constructions of Égyptien de tradition. This may also explain the fact that some patterns found are not perfectly in line with general typological expectations. Three miscellaneous sections deal with (i) a case of passive possession, (ii) cases of genitive constructions in which the analyses of the grammatical structure and the indications of the semantic structure via “phrase classifiers” do not go together, and (iii) a possessive construction that exhibits a curious hybrid of Earlier Egyptian grammar and Late Egyptian spelling habits. 0 Introduction In this article, I explore the usage of mainly two attributive possessive constructions, traditionally labeled as “Direct Genitive” construction and “Indirect Genitive” construction by Egyptologists. The Direct Genitive construction is a productive type of noun compound construction which is roughly comparable to German possessive noun compounds. The Indirect Genitive construction, on the other hand, is comparable to the English of-construction. In both constructions, the notion “genitive” refers not to a morphological case but to an attributive possessive function. More specifically, I mainly investigate these constructions in Egyptian texts written in Égyptien de tradition, i.e., in an artificial emulation of “ancient” language, which the Egyptians used for highly prestigious texts in the later 2 nd and 1 st millennia BCE. I start with an introduction to Ancient Egyptian chronolects (Section 1) and the phenomenon of Égyptien de tradition (Sections 1–2). A survey of possessive constructions in two Égyptien de tradition corpora (Section 3) sheds light on the knowledge about original Earlier Egyptian chronolects that the
39
Embed
“Genitive” possessive constructions in Égyptien de ...
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
“Genitive” possessive constructions in Égyptien de tradition:
Compound construction vs. of-construction
Daniel A. Werning (EXC “Topoi,” Humboldt-Universität zu Berlin)
Keywords: Egyptian, Égyptien de tradition, attributive possession, noun compound, of-construction,
emulation of ancient language, passive possession, animacy, alienability, genitive meaning, semantic map,
phrase classifiers.
Abstract
This article mainly explores the usage of two attributive possessive constructions that are traditionally
labeled the “Direct Genitive” and “Indirect Genitive” (in both cases, the notion “genitive” refers not to a
morphological case but to an attributive possessive function). The former is a productive type of noun
compound construction, the later is comparable to the English of-construction. The texts investigated are
written in Égyptien de tradition, i.e., in an artificial emulation of “ancient” language in the later 2nd and 1st
millennia BCE. Among other things, the article explores constructional as well as semantic influences on
the choice of one or the other construction in the Book of Caverns (13th century BCE): (i) influences of
constructional complexity of the two related nouns/noun phrases, (ii) the animacy of possessors, and (iii)
the alienability of the possessive relation. In the course of this, the statistical impact of cases of “agentive
possesseds” like, e.g., ‘ruler of the netherworld’ and “possessing possesseds” like, e.g., ‘possessor of a bier,’
is discussed, i.e., the impact of cases in which the possessed “controls” or possesses the possessor. Finally
(iv), the attested “genitive” meanings are mapped onto a semantic map of possessive relations. The
quantitative observations made are interpreted as the result of the application of a simple translation rule by
which the ancient authors transformed genitive constructions of their contemporary language varieties into
genitive constructions of Égyptien de tradition. This may also explain the fact that some patterns found are
not perfectly in line with general typological expectations. Three miscellaneous sections deal with (i) a case
of passive possession, (ii) cases of genitive constructions in which the analyses of the grammatical structure
and the indications of the semantic structure via “phrase classifiers” do not go together, and (iii) a
possessive construction that exhibits a curious hybrid of Earlier Egyptian grammar and Late Egyptian
spelling habits.
0 Introduction
In this article, I explore the usage of mainly two attributive possessive constructions, traditionally labeled as
“Direct Genitive” construction and “Indirect Genitive” construction by Egyptologists. The Direct Genitive
construction is a productive type of noun compound construction which is roughly comparable to German
possessive noun compounds. The Indirect Genitive construction, on the other hand, is comparable to the
English of-construction. In both constructions, the notion “genitive” refers not to a morphological case but
to an attributive possessive function. More specifically, I mainly investigate these constructions in Egyptian
texts written in Égyptien de tradition, i.e., in an artificial emulation of “ancient” language, which the
Egyptians used for highly prestigious texts in the later 2nd and 1st millennia BCE.
I start with an introduction to Ancient Egyptian chronolects (Section 1) and the phenomenon of
Égyptien de tradition (Sections 1–2). A survey of possessive constructions in two Égyptien de tradition
corpora (Section 3) sheds light on the knowledge about original Earlier Egyptian chronolects that the
Daniel A. Werning, "Genitive" possessive constructions in Égyptien de tradition: Compound construction vs. of-construction,to be
published in: Eitan Grossman & Stéphane Polis (eds.). Possession in Ancient Egyptian, Berlin: de Gruyter
authors of these texts obviously had. The main part of this article is a case study of full noun genitive
constructions in one particularly long text in early Égyptien de tradition: the Book of Caverns from the 13th
century BCE. After an introduction to the Direct and Indirect Genitive constructions in Earlier Egyptian as
opposed to Late Egyptian (Section 4.1), I proceed to analyze Direct and Indirect Genitive constructions in
the Book of Caverns (Section 4.2).
One section of this main part looks at possible influences of constructional complexity of the possessor
noun and the possessed noun on the choice of one or the other construction (Section 4.2.1). In the course of
this, I discuss the question whether the authors of texts in Égyptien de tradition conceptualized their “Direct
Genitive” as a Simple Juxtaposition Genitive Construction, rather than as a Direct Genitive proper, i.e., a
phonologically marked compound construction as in original Earlier Egyptian. Section 4.2.2 explores
possible semantic influences of the related nouns and/or the “possessive” meaning relation. We look at the
animacy of possessors, the alienability of the possessive relation, and I review the attested genitive
meanings on the basis of a semantic map of possessive relations that was proposed by Kiki Nikiforidou
(1991).
After the main part, three miscellaneous sections deal with a few interesting phenomena related to
possessive constructions in Caverns that are beyond the main scope of this article. The first section treats a
case of passive possession (Section 5.1). The second discusses cases of genitive constructions in which the
analyses of the grammatical structure and the indications of the semantic structure via “phrase classifiers”
do not go together (Section 5.2). The third section highlights a possessive construction that exhibits a
curious hybrid of Earlier Egyptian grammar and Late Egyptian spelling habits (Section 5.3).
A final paragraph contains a concluding summary (Section 6).
1 Egyptian chronolects and Égyptien de tradition
For highly prestigious texts such as, e.g., religious and official texts, the Egyptians always used a language
variety that was old-fashioned in some respects. Over the course of the later 2nd millennium BCE, the
difference between the language used for such texts and the contemporary spoken language had grown
considerably. For prestigious texts, the authors tried to emulate the “ancient” language that they found in
texts from past times, i.e., in texts from the 3rd and earlier 2nd millennium BCE. Egyptologist call these
emulations “Spätmittelägyptisch” (e.g., Karl Jansen-Winkeln), “Neo-Middle Egyptian” (Friedrich Junge),
or “Égyptien de tradition” (Pascal Vernus).1 Texts in Égyptien de tradition are attested until Roman times,
i.e., until the early 1st millennium CE (Figure 1).
1 See Jansen-Winkeln (1996), Junge (1984), Vernus (e.g., in L’Égyptologie en 1979), Stauder (2013b). I prefer the
term “Égyptien de tradition” to “Neo-Middle Egyptian” since the latter may lead the reader to the false
assumption that (Classical) Middle Egyptian was the only model for the emulation.
Daniel A. Werning, "Genitive" possessive constructions in Égyptien de tradition: Compound construction vs. of-construction,to be
published in: Eitan Grossman & Stéphane Polis (eds.). Possession in Ancient Egyptian, Berlin: de Gruyter
Fig. 1. Ancient Egyptian chronolects and text registers (based on Kammerzell 1998: Abb. 5 and Junge 1984:
Tab. 2; CC-BY)
We may assume that the ancient language was not spoken at home anymore at a certain point in time.
Instead, authors of such texts had to learn about the ancient language in private “schools,” from reading and
copying old texts and from what the senior scribes taught them. For a new text, they could partly copy and,
if necessary, adapt sentences and phrases that they found in original ancient texts. For the rest of the text as
well as for completely innovative texts, however, they needed to create new phrases and sentences in
Égyptien de tradition by themselves. Given that the authors’ primary spoken language, or rather language
varieties (L1), were considerably different from the “ancient” language varieties, we often find certain
influences of their L1 in Égyptien de tradition (L2) texts.2 On the one hand, this Égyptien de tradition
exemplifies constructions and lexical items that were saliently different in ancient texts as compared to
contemporary language. On the other hand, those constructions and lexical items that were considered
typical of the contemporary language were deliberately avoided. We may think of the process as a
translation of a mental text in L1 to a written text in L2 (see Section 2 below). In any event, as a result, texts
in Égyptien de tradition do not always perfectly match any authentic ancient chronolect like Old Egyptian
or Middle Egyptian from the 3rd and earlier 2nd millennia BCE. Furthermore, it is probably not the case that
Égyptien de tradition was taught and used in an unbroken, fixed scholarly tradition over more than one
thousand years of its attestation. To the contrary, we have to account for the fact that Égyptien de tradition
also changed and that it was seemingly even re-invented over and over again. It is therefore advisable to
study, rather than all texts in Égyptien de tradition at once, the Égyptiens de tradition of common “schools”
or even of individual texts.
2 Égyptien de tradition as mental translation product
In Section 1, I proposed thinking of the creation of an entirely new text in Égyptien de tradition as a
translation of a mental text in a contemporary used language variety of Egyptian (L1) to a written text in
Égyptien de tradition (L2). Egyptologists call the chronolects of the late 2nd and the 1st millennium BCE that
were comparatively close to the contemporary spoken language “Late Egyptian” and “Demotic,”
respectively (see Figure 1 above). The following text examples from the Book of Caverns,3 an early
Égyptien de tradition text from the 13th century BCE, clearly show the influence of the contemporary L1,
2 See Stauder (2013b), Werning (2011, I: 259–260), Werning (2013). 3 The text critical edition Werning (2011: II) is referred to by “Hb.” throughout this article.
Daniel A. Werning, "Genitive" possessive constructions in Égyptien de tradition: Compound construction vs. of-construction,to be
published in: Eitan Grossman & Stéphane Polis (eds.). Possession in Ancient Egyptian, Berlin: de Gruyter
i.e., Late Egyptian. For example, the constructions below match neither any original Earlier Egyptian
grammar, nor contemporary Late Egyptian grammar. Instead, they are artificial hybrids of features of both
grammars.4 Earlier Egyptian and Late Egyptian are different with respect to both the morphological set and
function of determiners, as well as with respect to the syntactic position of the determiner. In the Book of
Caverns, we occasionally find ancient determiner morphemes in the syntactic position of contemporary
determiners. The specific combinations of morphemes and syntactic positions are not attested in any
original chronolect.
In example (1), we find an Earlier Egyptian demonstrative morpheme in the position of the Late
Egyptian demonstrative (the “*” before Egyptian text examples marks reconstructed phrases).
(1) L2 = Ég. de trad.
(13th c. BCE) (Hb. 67.7, tO)
m= tn Hry.t
in= DEM.PROX:F oven:F
Class. Middle Eg.
(20th c. BCE) * (reconstruction)5
m= Hry.t =tn
in= oven:F =DEM.PROX:F
L1 = Late Eg.
(13th c. BCE) * (reconstruction)
m= t#~ Hry(t)
in= DEM:F oven(F)
‘into this oven’
In (2) and (3), we find an Earlier Egyptian demonstrative morpheme with the function of a determinative in
the position of the Late Egyptian definite article or possessive article, respectively. Original written Earlier
Egyptian did not yet employ any morpheme in the function of a determinative.
4 Werning (2011, I: 186–191, §§112–119). 5 Note that the lemma Hry.t ‘oven’ is not clearly attested before the middle of the 2nd millennium BCE (Werning
2011, I: 250; for the debated dating of Kheti, see now Stauder (2013a: 468–476).
Daniel A. Werning, "Genitive" possessive constructions in Égyptien de tradition: Compound construction vs. of-construction,to be
published in: Eitan Grossman & Stéphane Polis (eds.). Possession in Ancient Egyptian, Berlin: de Gruyter
(3) L2 = Ég. de trad.
(13th c. BCE) (Hb. 90.17, RVI)
m= pn= HD.y-j
INSTR= DETV:M.SG= bright\PART.DISTR[M.SG]-1SG
Class. Middle Eg.
(20th c. BCE)
* (reconstruction)
m= HD-j
INSTR= bright\PART[M.SG]-1SG
L1 = Late Eg.
(13th c. BCE)
* (reconstruction)
m= p{#}y-j HD
INSTR= DEF.POSS:M.SG-1SG bright
‘with my bright one (i.e., the sun disc)’
In (3), the position of the personal suffix pronoun that indicates the possessor is “ancient” insofar as it is
attached to the possessed, as in Earlier Egyptian, and not to the article as in Late Egyptian. The resulting
construction is alien to Earlier Egyptian as well as to Late Egyptian.6
In Section 4, we explore comparable phenomena in two competing attributive possession constructions,
called the “Direct Genitive” and the “Indirect Genitive” by Egyptologists. Before doing so, however, I
survey the set of attested possessive constructions in some Égyptien de tradition texts.
3 Survey of possessive constructions in two Égyptien de tradition corpora
In Section 1 and 2, we saw that Égyptien de tradition is an artificial emulation of Earlier Egyptian
grammars and that it occasionally exhibits influences from the authors’ contemporary spoken language
varieties. In order to get an impression of which Earlier Egyptian possessive constructions are attested in
Égyptien de tradition texts, Table 6 in the appendix enumerates constructions and their attestation in two
such corpora. The first is the “Book of Caverns” (13th century BCE; Werning 2011). The second is a corpus
of texts from the Third Intermediate Period (henceforth TIP; 11th–8th century BCE; Jansen-Winkeln 1996).
Interpreting Table 6, we may summarize that most Earlier Egyptian constructions that are attested more
than only exceptionally in Earlier Egyptian texts are attested in Caverns as well. Only two comparatively
rare attributive constructions are missing,7 as well as two specific existential clause patterns8. Furthermore,
constructions commonly or exclusively attested in Late Egyptian are avoided.9 This may also explain the
mentioned absence of the two missing attributive constructions and one of the missing existential clauses,
similar forms of which are also part of Late Egyptian grammar.10 The basic trend also seems to hold in
Égyptien de tradition texts from the Third Intermediate Period. In this corpus, however, a few Late
6 For comparable and rarely attested constructions in Late Egyptian (*p#=HD-j DEF:M.SG=bright-1SG or *p{#}y HD-j
DEM:M.SG bright-1SG), see Winand (this volume). 7 The constructions with deictic adverb jr-~ ‘thereof’ and n.~-...=jm.y ‘of mine, of yours, ...’ (see Table 6 in the
appendix). 8 The affirmative existential clause [jw wn N-PRON] and the negated pattern [nn N-PRON] without wn (see Table 6
in the appendix). 9 E.g., the possessive article (p#y-, t#y-, n#y-), the n(~).sA clause, the [N sw(t)] construction, the possessive pronouns
p(#).n(~) and tj.n(~)t, and the wn=md~ clause (see Table 6 in the appendix). 10 Earlier Eg. jr-~ and n(.~)-...=jm.y survive in Late Eg. as uninflected jr~ and n(~)-...=jmy, respectively. Earlier Eg.
[jw wn N-PRON] is comparable to Late Eg. [wn N ...] (see Table 6 in the appendix).
Daniel A. Werning, "Genitive" possessive constructions in Égyptien de tradition: Compound construction vs. of-construction,to be
published in: Eitan Grossman & Stéphane Polis (eds.). Possession in Ancient Egyptian, Berlin: de Gruyter
Egyptian constructions are also attested.11 As opposed to Caverns, one “ancient” Earlier Egyptian
existential clause is not attested.12
Altogether, we see that the authors that composed texts in Égyptien de tradition were remarkably well
informed about Earlier Egyptian, and that they were very aware of typical differences between Earlier
Egyptian and Late Egyptian. Therefore, when we explore possible differences between Égyptien de
tradition and original Earlier Egyptian in a specific complementary pair of constructions in the next section
(Section 4), namely the “Direct Genitive” and the “Indirect Genitive,” we may expect rather subtle
differences, i.e., differences that were not salient enough to be necessarily noticed or differences triggered
by contextual criteria too deeply entrenched in the spoken language of the authors to escape.
4 Case study: full noun genitive constructions in the Book of Caverns
4.1 Full noun genitive constructions in Earlier and Late Egyptian
For the expression of a possessive relation between two full nouns, Egyptian of the 3rd and 2nd millennia
BCE basically had two different attributive constructions that Egyptologists call “Direct Genitive” and
“Indirect Genitive,” respectively.13 The Indirect Genitive resembles the English construction [NPOSSED of
NPOSSOR], with the difference that the attributive/“genitive” connector exhibits gender/number agreement14
with the head of the phrase, the possessed noun, in Earlier Egyptian.15 The agreement, however, became
increasingly optional and was eventually lost in Late Egyptian (Table 1).
11 E.g., the possessive article (p{#}y-, t{#}y-, n{#}y-), the n(~).sA clause, and n(~)-...=jmy (see Table 6 in the appendix). 12 The non-existence clause with wn.t (see Table 6 in the appendix). 13 Loprieno (1995: 56–57), Allen (2013: 71); Schenkel (1962), Kammerzell (2000: 101–104); Jansen-Winkeln
(2000). 14 On the agreement of the connector in genitive constructions as a possible typological criterion, see Serzisko (2004:
1093–1094). 15 According to the Egyptological opinio communis, the genitival connector n(~) / n(~)-w / n(~)-t ‘GEN, of,’ regularly
spelled simply n, nw, or n-t, respectively, was probably etymologically an adjectival derivation of the “dative”
preposition n ‘DAT, to, for’ (e.g., Malaise & Winand 1999: §92, Schenkel 2012: §5.2.3, Allen 2013: 74). Earlier
Egyptian had a productive means to derive an adjective from a noun or preposition with an adjectivizing suffix -~,
which Egyptologist and Semitists call a “nisbe” derivation: n-~ / n-~-w / n-~-t. Since the nisbe morpheme -~ itself
was most likely vocalic */iː/ (Werning 2011, I: 139–141, §§56–57; Schultheiß 2014) and there was also no
immediate need for a mater lectionis (cf. Werning forthcoming), it was normally not reflected in the hieroglyphic
script (n, nw, n-t).
Daniel A. Werning, "Genitive" possessive constructions in Égyptien de tradition: Compound construction vs. of-construction,to be
published in: Eitan Grossman & Stéphane Polis (eds.). Possession in Ancient Egyptian, Berlin: de Gruyter
Table 1. The “Indirect Genitive” construction in Old Egyptian and Late Egyptian
Some Egyptologists take this head marking construction to be an old compound construction that was not
productive anymore already in the 2nd millennium BCE.17
It seems, however, that the stress in the Direct Genitive construction was usually on the second noun,
i.e., the possessor, and that the first noun appeared in a phonetically somewhat reduced form, which
Egyptologists and Semitists call status constructus,18 for example:
16 Occasionally, we still find the spellings n-t and nw in Late Egyptian. Some of the instances seem to
retain or emulate Earlier Egyptian grammar. In other instances, however, the spellings seem to refer to the same
spoken morpheme as the spelling n (see Erman 1933: §§213–215, Junge 1999: 64). 17 Cf. Edel (1955/1964: §320), Schenkel (1990: 81–86, with reference to G. Fecht), Loprieno (1995: 57),
Kammerzell (2000: 101–102); but cf. Allen (2013: 71). Roman Gundacker is currently working on a book on
account that simplifies the actually attested variation to a certain degree, Kammerzell (2000: 101–104) proposed
that the distribution of Direct Genitives vs. Indirect Genitives corresponded to an alienability split and that it
already developed in the late 3rd millennium BCE. While this might come close to the attested variation for the
specific case of pronominal possessors (cf. the contributions of Gardiner, Sojic, and Winand in this volume), the
distribution for the case of nominal possessors seems more complex. 24 Jansen-Winkeln (2000: 37); cf. also few remarks by Shisha-Halevy (2007: 239–240). For comparable statistical
studies on the English s-Genitive vs. of-Construction, see Rosenbach (2003) and Altenberg (1982). 25 Jansen-Winkeln (2000: §§6, 8). 26 Jansen-Winkeln (2000: 30, 31). 27 Cf. Schenkel (1962); Jansen-Winkeln (2000: §§5, 7); recently, e.g., Borghouts (2010, I: §§83a, 85.c.1).
Daniel A. Werning, "Genitive" possessive constructions in Égyptien de tradition: Compound construction vs. of-construction,to be
published in: Eitan Grossman & Stéphane Polis (eds.). Possession in Ancient Egyptian, Berlin: de Gruyter
in ‘a statue of stone’ or ‘a man of honor,’ the Indirect Genitive is obligatory.28 He points out that, in these
cases, the distribution of the actual possessive semantics of the relation contradicts the distribution of the
possessive semantics of the possessive construction: While, grammatically, the “statue” is in the position of
the possessed, semantically it is rather something like a possessor (of a quality). With a certain type of
nouns, however, we also find comparable meanings with Direct Genitives, e.g.,29
(10) Erman & Grapow (1962–1931, II: 228,12–13),
nb==m#o.t BBAW, DZA (23.762.840–850)
lord(M)==truth:F
‘possessor of truth; one who is truthful.’
Below, I suggest classifying examples like this as special cases in which lexical influences overrule the
constructional meaning (Section 4.2.2, meaning (m2b) “possessing possesseds”). Leaving these cases aside,
we may still adhere to the hypothesis that qualitative possession is not normally expressed by Direct
Genitives.
It is fruitful to relate this hypothesis to a semantic map of possessive meaning developed by Nikiforidou
(1991). She suggested that the set of observable meanings of Indo-European possessive constructions be
analyzed as a network of related meanings that are more or less remote from prototypical possession
(Figure 2; see also Section 4.2.2 below).
Fig. 2. Semantic map of IE possessive constructions after Nikiforidou (1991: appendix)
We see that genitive constructions with qualities in the grammatical slot of the possessor, i.e., Thing
constituted | Constituent material and Holder of an attribute | (Distinctive) property, are comparatively
remote from the prototypical possessive meaning of Possessions | Possessor. In the light of this research, I
28 See the meaning (f) Thing constituted | Constituent material in Section 4.2.2. Schenkel (1962: 60) already hints at
the impossibility of expressing the possession type with a material as the possessor with a Direct Genitive. 29 Compare also Eyre (this volume).
Daniel A. Werning, "Genitive" possessive constructions in Égyptien de tradition: Compound construction vs. of-construction,to be
published in: Eitan Grossman & Stéphane Polis (eds.). Possession in Ancient Egyptian, Berlin: de Gruyter
propose that the above-mentioned observations on genitive constructions in Earlier Egyptian can be
translated into a more general hypothesis on semantic influences on the choice of the Direct or Indirect
Genitive construction:
iii) Semantic influences: Less prototypical possessive relations prefer the Indirect Genitive to the Direct
Genitive.
In the following section, we will evaluate possible influences on the choice of the Direct or the Indirect
Genitive based on observations about frequencies in one specific, comparatively long text in Égyptien de
tradition, namely the Book of Caverns.
4.2 Full noun genitive constructions in the Book of Caverns
The Book of Caverns was composed in the 13th century BCE.30 It contains approximately 11,500 words and
526 instances of genitive constructions (hereby, literal repetitions in litanies are not counted). The following
statistics are based on raw data that are available from xxx ### ONLINE REPOSITORY TO BE
INSERTED ### xxx.
The contemporary spoken language and the contemporary language of private letters and of literature
was Late Egyptian (cf. Figure 1, above). The language of Caverns is an ancient scholarly emulation of
earlier language, i.e., an Égyptien de tradition (Section 1). We may think of the creation of the largely new
text of Caverns as a mental translation of their Late Egyptian (L1) thoughts into Égyptien de tradition
(Section 2). Generally speaking, the Égyptien de tradition of Caverns is quite similar to original Earlier
Egyptian. There are, however, also certain influences from the L1 of the authors, i.e., Late Egyptian.31 In
this section, I describe the usage of genitive constructions with two or more full nouns and try to spot
possible Late Egyptian influences.
First, we find many more instances of Direct Genitives in Caverns than instances on Indirect Genitives
(Table 3).
Table 3. Relation of Direct vs. Indirect Genitives in the Book of Caverns
Direct Genitive
NPOSSED==NPOSSOR
Indirect Genitive
NPOSSED n(~)/n(~)-w/n(~)-t NPOSSOR
Percent of instances ca. 80% ca. 20%
Given that the Direct Genitive became increasingly rare in the non-artificial language over the course of the
2nd millennium BCE (Section 4.1), this seems to be a remarkable high frequency of the Direct Genitives.
Regarding the Indirect Genitive construction, Earlier Egyptian and Late Egyptian are fundamentally
different with respect to the agreement of the “genitive” connector. We find gender/number agreement of
the genitive connector with the head noun in Earlier Egyptian, but no agreement in Late Egyptian (Table 1
‘the enemies of the one who is at the head of the netherworld.’
36 At least Xft(~)-w==(W)sjr ‘the enemies of Osiris’ is very probably a Direct Genitive compound. Compare the
position of the enclitic demonstrative copula pw in the following Nominal Sentence:
Xft(~)-w==Ro=pw [opponent(M)-PL==Ra(M)]=DEM ‘These/they are the-enemies-of-Ra’ (Hb. 68.6, Pet). This
copula is normally attached to the first compact word or phrase. 37 The word Xft~ ‘the enemy, the opponent,’ originally also a nominalized de-prepositional adjective/nisbe Xft-~
(opposite-ADJZ[M.SG]) ‘the one who is opposed,’ on the other hand, is very probably already a fully lexicalized
noun (see footnote 36). 38 For the meaning ‘one who dwells in,’ see Werning (2011, II: 506). 39 For the meaning ‘one close to, one at,’ see Werning (2012: §§4.1–4.2). 40 For the meaning ‘one who is (in the form of),’ see Werning (2011, II: 478).
Daniel A. Werning, "Genitive" possessive constructions in Égyptien de tradition: Compound construction vs. of-construction,to be
published in: Eitan Grossman & Stéphane Polis (eds.). Possession in Ancient Egyptian, Berlin: de Gruyter
In this particular case, however, analysis (b) is probably less likely. In the written form, the phrase Xnt.~
dw#.t is spelled with a phrase classifier [GOD] which refers to the phrase Xnt~ dw#.t as a whole.41 This
indicates that the authors understood this pair of words as a combined expression ‘chief–of–the–
netherworld.’ However, since these cases with de-prepositional nisbes are not clear examples for genitive
constructions, they are left out of the statistics in this article.
The following example (18) is a comparable case that is also difficult to analyze since a nominalized
adjective/nisbe derived from the compound preposition m=Xt ‘behind, after, in the following’ is involved.
Traditionally, Egyptologists assume that the connection between the relational noun, Xt in this case, and the
following noun is a Direct Genitive connection (m=Xt==N). It is unclear, however, whether this necessarily
holds true for the de-prepositional nisbe jm.~ Xt ‘one who is behind, one following, follower’ as well.
(18) Caverns
(13th c. BCE)
(Hb. 4.1–2, RVI)
j psD.t n(~)-t nTr-w
o! ennead:F[SG] of-F god(M)-PL
jm(.~)-t Xt=?=k#==jmn.t
in:ADJZ-F following?(M)=?=bull(M)[SG]==west:F[SG]
‘O divine group of gods,
the ones in the following of the “bull” of the west!’42
To conclude, the Égyptien de tradition in Caverns exemplifies 14 clear and another 23 debatable instances
of constructions that are either to be analyzed as nested Direct Genitive Constructions [N==[N==N]] ‘the N
of the N of N’ or, alternatively, as constructions that include at least one Simple Juxtaposition Genitive
construction [N N] ‘the N of N.’ However, both constructions are, as far as I can see, alien to both Earlier
and Late Egyptian. I propose that these constructions are the result of a simple rule for the translation of
genitive constructions that were mentally phrased in the author’s contemporary chronolect Late Egyptian
(L1) into Égyptien de tradition (L2). The rule was, evidently, to (i) delete all articles and (ii) either
exchange the Late Egyptian uninflected connector n(~) ‘of’ in the Indirect Genitive construction by an
inflected version of it, i.e., n(~)-ø, n(~)-t, or n(~)-w, respectively, or to simply leave the connector out (see
Table 5). It is unclear whether they conceptualized the latter case, i.e., two directly following nouns, as a
phonetically marked Direct Genitive compound or as a Simple Juxtaposition Genitive Construction.
41 A phrase classifier refers to the combined meaning of a whole phrase or productively derived word form and
usually stands at the end of it. For the notion and the phenomenon of “phrase classifiers,” see Werning (2009: 5),
Werning (2011, I: 106–107, §12 with footnote 62), Werning (in preparation). For a largely similar approach,
cf. Lincke (2011: §2.3), Lincke & Kammerzell (2012: §3.3.2–3). Note, however, that compound semantics need
not necessarily imply compound morpho-syntax (cf. Haspelmath 1998). 42 Note that the spelling of jm(.~)-t Xt with the phrase classifiers [GOD][PLURAL] indicates that jm(.~)-t Xt was
understood as a combined expression ‘the–ones–in–the–following’ and also that it was not felt as simply an
adjectival attribute to psD.t ‘ennead’ but as a full noun phrase in apposition. Cf. a similar example in Hb. 101.9.
The meaning of the relational noun Xt alone is not entirely clear (cf. Erman & Grapow 1926–1931, III: 342–347).
Daniel A. Werning, "Genitive" possessive constructions in Égyptien de tradition: Compound construction vs. of-construction,to be
published in: Eitan Grossman & Stéphane Polis (eds.). Possession in Ancient Egyptian, Berlin: de Gruyter
Table 5. Translation process model of genitive constructions in Égyptien de tradition
Late Egyptian (L1), bipartite (ART=)N n(~)= (ART=)N
(ART=)N== N
Translation rules: ↓delete ART ↓inflect,
or delete
↓delete ART
Égyptien de tradition N=?= N
N n(~)-ø/t/w N
Late Egyptian (L1), tripartite (ART=)N n(~)= (ART=)N n(~)= (ART=)N
(ART=)N n(~)= (ART=)N== N
(ART=)N== N n(~)= (ART=)N
Translation rules: ↓delete ART ↓inflect,
or delete
↓delete ART ↓inflect,
or delete
↓delete ART
Égyptien de tradition N=?= N=?= N
N n(~)-ø/t/w N=?= N
N n(~)-ø/t/w N n(~)-ø/t/w N
N=?= N n(~)-ø/t/w N
In two interesting instances of genitive constructions, the authors of
Caverns did not apply the general rule to simply delete all Late
Egyptian (L1) articles (example 19). Instead, they exchanged them
with an Earlier Egyptian demonstrative morpheme (cf. examples 1–
3 above). The resulting construction is definitely an Égyptien de
tradition hybrid that is alien to both Earlier and Late Egyptian.
Again, it is unclear whether the resulting construction was
conceptualized as a Direct Genitive construction or as a Simple
Juxtaposition Genitive Construction.
Fig. 3. The Mysterious One with Ra
on her hand (Book of Caverns, scene
L; Werning 2011, II: 234)
(19) Caverns
(13th c. BCE)
(Hb. 60.5, tO; sim. 60.12)
Hr o(w)-w~=?=tn=St{#}y-t
on arm&hand(M)-DU=?=DETV:F=mysterious-F
L1 = Late Eg.
(13th c. BCE) * (reconstruction)
Hr= o(w).w~ n(~)= t#=St{#}y(t)
on= arm&hand(M):DU of= ART:F=mysterious(F)
Class. Middle Eg.
(20th c. BCE) * (reconstruction)
Hr o(w)-w~==St#-t
on arm&hand(M)-DU==mysterious-F
‘on the hands of The Mysterious One’ (cf. Figure 3)
Daniel A. Werning, "Genitive" possessive constructions in Égyptien de tradition: Compound construction vs. of-construction,to be
published in: Eitan Grossman & Stéphane Polis (eds.). Possession in Ancient Egyptian, Berlin: de Gruyter
4.2.2 Possible semantic influences
Generally speaking, possible semantic influences on the choice of either the Indirect Genitive or the Direct
Genitive constructions may come from (i) semantic features of the possessor, (ii) semantic features of the
possessed, and/or (iii) the semantic relation between the two. In this section, we first look at the animacy of
possessors in the Book of Caverns. Afterwards, we investigate the degree of the alienability of the attested
relations. Finally, I review the attested semantic relations based on a semantic map of possessive meanings
that was suggested by Nikiforidou (1991).
The animacy of the possessor
Research on possession often claims that “prototypical instances of possession involve human possessors,
and, less typically, non-human animate possessors.”43 In Caverns, we find that genitive possessors are
altogether nearly as often inanimate as animate (Figure 4).
Fig. 4. Animacy of genitival possessors in the Book of Caverns (CC-BY)
Furthermore, Figure 4 shows that the Direct Genitive [N==N] and the Indirect Genitive [N n(~) N] behave
differently with respect to animacy. While both animate and inanimate possessors prefer the Direct Genitive,
it is surprisingly44 not the animate, but the inanimate possessors that prefer the Direct Genitive even more.
In the Indirect Genitive, inanimate possessors are remarkably rare.
The results, however, are influenced by the statistical effect of special cases like, for example, ‘ruler of
N,’ ‘lord of N,’ ‘king of N,’ or ‘possessor of N’ in which the grammatical slot of the possessed is occupied
by an agent or a (semantic) possessor and, consequently the slot of the possessor is occupied by a
patient/theme or a (semantic) possessed. I tentatively label cases like these “agentive possesseds” or
“possessing possesseds,” respectively (see meaning (m2) below). If we delete these cases from our set, we
find, indeed, that possessors are more often animate than inanimate, and a relation between the animacy of
the possessor and the choice of one or the other genitive construction is not that clear (Figure 5).
43 Heine (1997: 21, cf. 5, 9). 44 Cf. the opposite phenomenon in English (Rosenbach 2003: §4.3).
Daniel A. Werning, "Genitive" possessive constructions in Égyptien de tradition: Compound construction vs. of-construction,to be
published in: Eitan Grossman & Stéphane Polis (eds.). Possession in Ancient Egyptian, Berlin: de Gruyter
Fig. 5. Animacy of genitival possessors in Caverns without cases of agentive and possessing possesseds
(CC-BY)
The alienability of the possessive relation
In an evaluation of the relation of alienable possession and head marking in Ancient Egyptian, Kammerzell
(2000: 102) claimed that in Ancient Egyptian around 2000 BCE, i.e., in Middle Egyptian, “[p]ossessive
expressions with alienable possessed are no longer built productively by means of compounds (i.e., Direct
Genitives, D.W.).” Due to the lack of actual quantitative research on this topic, however, we can only take
this as the educated intuition of an experienced Egyptological linguist. Assuming that Kammerzell’s
intuition is accurate, we may expect that authors who composed texts even later in the 2nd millennium
expressed alienable possessive relations preferably with the Indirect Genitive. In the Égyptien de tradition
of Caverns, however, we find exactly the opposite behavior. The authors did not only generally prefer the
Direct Genitive to the Indirect Genitive (Table 4 above), but, in cases of alienable possession, they used the
Indirect Genitive not in a higher but rather in a lower percentage of cases than in the cases of inalienable
possession (Figure 6).
Fig. 6. Alienability of genitival possessive relations in the Book of Caverns (CC-BY)
Actually, the data in Caverns contradict possible expectations suggested by iconicity principles in language.
The iconicity principle predicts that the closer semantic relation, i.e., inalienable possession, tends to be
expressed by the more compact construction, i.e., the Direct Genitive, while the more loose relation, i.e.,
alienable possession, tends to the less compact construction, i.e., the Indirect Genitive.45 Rather than taking
the opposite tendency in Caverns as a counterexample against iconicity principles, I suspect that the
45 Cf. Rosenbach (2003: 389, 398), Pusch (2001: §4.2.1), with reference to the work of John Haiman.
Daniel A. Werning, "Genitive" possessive constructions in Égyptien de tradition: Compound construction vs. of-construction,to be
published in: Eitan Grossman & Stéphane Polis (eds.). Possession in Ancient Egyptian, Berlin: de Gruyter
unexpected data in Caverns testify for the artificial nature of its language. Generally, it seems that we
cannot claim that we find an alienability split in Caverns, at least not an obvious one.
A semantic map of possessive relations
In an article on genitives, Nikiforidou (1991) suggests analyzing the set of observable meanings of
possessive constructions as a network of metaphorically related meanings. Her main result is a
corresponding semantic map of genitive meanings (Figure 2 above). In this section, we explore the genitive
meanings that we find in Caverns and map them on a rearranged and slightly expanded version of
Nikiforidou’s semantic map (Figure 7 below).
The following list exemplifies genitive meanings that we find in genitive constructions in Caverns. I
tried to categorize the meanings based on Nikiforidou’s classification (meanings (a)–(m) below). Eventually,
I specified the meanings more precisely, e.g., the meanings (c1)/(c2), (k1)/(k2), (m1)/(m2). However, some
cases are ambiguous and some do not clearly fit one of her classes, e.g., those mentioned in (c1a2) and (c2).
Two meanings seem to be missing in her map: Entity | Identified and Agent | Patient/Theme (see (n) and
(m2) below). In the case of the Entity | Identified, it is an identification that is established between the
possessor and the possessed. The other missing cases that may be categorized as cases of Agent |
Patient/Theme are cases in which the slot of the constructional possessed is filled by an actor or (semantic)
possessor, while the slot of the constructional possessor is filled by a patient, theme, experiencer, or
(semantic) possessed. In these cases it is the constructional possessed that “controls” or “possesses” the
constructional possessor (and not vice versa). Note that in the case of (m2b) Possessor | Possessed, the
meanings of the construction slots contradict the possessive meaning of the lexemes that occupy the slots
(constructional possessed—semantic possessor, constructional possessor—semantic possessed). I
tentatively label these special cases as cases of “agentive possesseds” and “possessing possesseds,”
respectively (see Section 4.2.2 above).
The following list gives examples for genitive meanings in the Book of Caverns:46
‘The ruler of the netherworld’ (Hq(#)==dw#.t), ‘the care-taker of this following’ (nDt.~ n(~) jm(~)-w Xt-f),
‘the heir of Osiris’ (j(w)ow==(W)sjr), ‘the enemies of Osiris’ (Xft(~)-w n(~)-w (W)sjr), ‘enemies of Ra’
(Xft(~)-w==Ro); ‘lord of the west’ (nb==jmn.t), ‘king of the west’ (nzw==jmn.t); ‘“great” of the west,
chief of the west’ (o#==jmn.t); (cases with de-prepositional nisbes, which are excluded from the statistics,
see the discussion above:) ‘chief of the netherworld’ (Xnt-~=?=dw#.t, see example 17, above), ‘master of
the lamenting ones’ (Hr(-~)=?=j#kb-(y)w).
(m2b) Possessor | Possessed (“possessing possesseds,” compare with (a))
‘“lord” of (his) head, possessor of (his) head’ (nb==D#D#), ‘“lord” of a bier, possessor of a bier’
(nb==Dry.t), ‘“lord” of a tassel, possessor of a tassel’ (nb==monX.t).
48 The difficulty to always clearly separate between Action | Theme/Patient and Experience | Experiencer is also
addressed by Nikiforidou (1991: 180). 49 Nikiforidou (1991: 153) mentions a case like this, namely “addresser of Sokrates,” only in an introductory list.
Daniel A. Werning, "Genitive" possessive constructions in Égyptien de tradition: Compound construction vs. of-construction,to be
published in: Eitan Grossman & Stéphane Polis (eds.). Possession in Ancient Egyptian, Berlin: de Gruyter
‘In your form of(/which is) Osiris-Orion’ (m= jrw-k n(~) (W)sjr c#H, see example 11, above), ‘in his
mysterious [for]m of(/which is) the one who wears a headscarf’ (m= <jr>w-f St# n(~) ofnt-~), ‘in your name
(of/)which is sčpyw’ (m=rn-Tn n(~) sTp-yw).
As for the additional meaning (n) Entity | Identified, I tentatively suggest understanding it as a metaphorical
expansion of the meaning (f) Thing constituted | Constituent material (see Figure 7). Support for this
hypothesis may come from the observation of prepositional meaning in Egyptian. The Earlier Egyptian
preposition m also covers both meanings ‘out of’ (material) and ‘as, being’ (essive).50
Figure 7. Semantic space of genitive constructions, slightly modified variant based on Nikiforidou (1991:
appendix) (CC-BY-SA)
The following diagrams (Figures 8–9) map the semantic spaces covered by the Direct Genitive (N==N,
compound construction) and the Indirect Genitive (N n(~) N, of-construction) in Caverns, respectively.51
50 Besides the meanings ‘in (inessive), from (ablative), with (instrumental).’ See Erman & Grapow (1926–1931, II:
1–2); for the spatial meanings of m, see Werning (2012: §5.2). 51 The numbers in brackets that are added to the meaning Part | Whole are cases of the meaning (c1a2) Divine
manifestation | Divinity. For those in brackets added to the meaning Holder of an attribute | (Distinctive) property,
different reading of the passages are possible. The meaning Effect | Cause is not attested in genitive constructions
in Caverns.
Daniel A. Werning, "Genitive" possessive constructions in Égyptien de tradition: Compound construction vs. of-construction,to be
published in: Eitan Grossman & Stéphane Polis (eds.). Possession in Ancient Egyptian, Berlin: de Gruyter
Figure 8. Semantic space of Direct Genitive constructions (N==N) in the Book of Caverns (CC-BY)
Figure 9. Semantic space of the Indirect Genitive construction (N n(~) N) in the Book of Caverns (CC-BY)
Evaluation of the sematic maps
Since the language of Caverns is obviously a mixture of sophisticated emulations of older chronolects and
influences of the contemporary spoken language, i.e., an artificial hybrid (Section 2, above), we should be
prepared for the eventuality that typological expectations are not necessarily met.
In concord with the concept of semantic maps, we expect the attested meanings of a genitive gram to
cover one continuous area on the map.52 Looking at the semantic maps (Figures 8–9, above), we see that the
semantic spaces of the two genitive constructions in Caverns largely cover contiguous areas despite the
artificial nature of its language. There are, however, some cases of Direct Genitive constructions that I
classified as cases of the meaning (g) Holder of an attribute | (Distinctive) property (see above) that seem to
form an island on the semantic map (Figure 8).
Earlier, I put forth the hypothesis, based on earlier claims concerning genitive constructions in Earlier
Egyptian, that less prototypical possessive relations prefer the Indirect Genitive to the Direct Genitive
(Section 4.1 above). This expectation seems to be met only partially. As expected, two comparatively
52 Haspelmath (2003: especially 217), Grossman and Polis (2012: 183–187, especially 186).
Daniel A. Werning, "Genitive" possessive constructions in Égyptien de tradition: Compound construction vs. of-construction,to be
published in: Eitan Grossman & Stéphane Polis (eds.). Possession in Ancient Egyptian, Berlin: de Gruyter
peripheral meanings, namely (f) Thing constituted | Constituent material and (n) Entity | Identified, are
always expressed with and only with the Indirect Genitive construction. On the other hand, the peripheral
meaning (g) Holder of an attribute | (Distinctive) property is attested with both constructions (given my
classifications of the examples are acceptable). And another similarly peripheral meaning, namely (h) Entity
compared | Standard of comparison, is attested with the Direct Genitive construction only. Given the low
number of instances of this meaning, however, the latter observation may simply be attributed to statistical
effects (this is why there are dotted lines in Figure 9). On average, we expect only 1 of 5 possessive
relations to be expressed with an Indirect Genitive (Table 3, above). The same statistical effect may also
explain why the meaning (k) Experiences/Products | Experiencer/Agent is not attested in Indirect Genitive
construction.
As far as the meaning (b) Kin | Person is concerned, on the other hand, we could have expected one or
two of the nine cases to be expressed with an Indirect Genitive. Therefore, the gap seems indicative. Taking
a closer look, however, we find that all nine instances of kinship relations in Caverns involve the lexeme
z# ‘son’ as the possessed (examples 20–23):
(20) z#==nTr son(M)==god(M) ‘son of a god’ (e.g., Hb. 48.11, RVI),
(21) z#==(W)sjr son(M)==Osiris(M) ‘son of Osiris’ (Hb. 101.36, RVI),
(22) z#==t# son(M)==earth(M) ‘son of the earth’ (Hb. 2.22, tO, RVI),
(23) z#-w==t# son(M)-PL==earth(M) ‘sons of the earth’ (e.g., Hb. 15.21, tO).
The lexeme z# ‘son,’ however, is a very frequent possessed noun in Egyptian, used in affiliations of the
pattern A z#==B ‘A, son of B’ in both, Earlier and Late Egyptian.53 In certain common phrases, z# even
made it into the 1st millennium CE, notably also in (21/24):54
(24) Old Eg. z#==(W)sjr son(M)==Osiris(M) */ʦiʀuːˈsiːrV/ ‘son of Osiris’
> Coptic ⲥⲓⲟⲩⲥⲓⲣⲉ siousire */siːuːˈsiːrə/ ‘son of Osiris.’
Moreover, the compound z#==t# (examples 22, 25) is among the few fixed old compounds of the type with
a stress on the first part that made it into Coptic times (see Section 4.1, example 4 above):55
(25) Old Eg. z#==t# son(M)==earth(M) */ˈʦiʀtaʀ/ ‘son of the earth (i.e. snake)’
> Coptic ⲥⲓⲧⲉ site */ˈsiːtə/ ‘basilisk.’
It was obviously common for z# ‘son’ to be used in a compound construction in all phases of the Egyptian
language. The authors of Caverns did not deviate from this usage in their Égyptien de tradition.
Due to the lack of quantitative studies on Earlier Egyptian as well as Late Egyptian, it is difficult to
further evaluate the semantic spaces of the two genitive constructions in Caverns. In cases in which Earlier
Egyptian and Late Egyptian chronolects seem to exhibit the same usage, the authors at least did not deviate
from it (e.g., in the use of z# ‘son’ and in the use of the Indirect Genitive construction for Thing constituted |
Constituent material relations). The general preference of the Direct Genitive to the Indirect Genitive in a
53 Compare also the traditional usage of z# with directly suffixed personal pronouns (see Winand, this volume). 54 Cf. Vycichl (1983: 182), Westendorf (1992: 535). 55 Westendorf ([1965] 1992: 198).
Daniel A. Werning, "Genitive" possessive constructions in Égyptien de tradition: Compound construction vs. of-construction,to be
published in: Eitan Grossman & Stéphane Polis (eds.). Possession in Ancient Egyptian, Berlin: de Gruyter
ratio of 4:1 in Caverns (see Table 3, above), however, is a hypercorrection toward an unnaturally frequent
use of the Direct Genitive compared to not only Late Egyptian, but very probably also compared to original
Earlier Egyptian chronolects.
5 Some remarkable possessive constructions in the Book of Caverns
5.1 Passive possession
The Book of Caverns exemplifies many examples of what is called “passive possession” in typological
literature.56 In each case, the possessed noun is kk.w ‘darkness’ and the possessor noun in a personal suffix
pronoun, as in (26).
(26) (Hb. 62.12, tO)
dr-j kk.w-Tn
drive_away:SBJV-1SG darkness:M.COLL-2PL
‘I will drive away the darkness around you (literally: your darkness).’
5.2 A puzzle: grammatical phrases vs. phrase classifiers
An often attested phrase in Caverns is jr(.~)-w o# n(~) N (the_one_at-M.PL door(M)[SG] of N), which most
naturally translates into English as ‘the doorkeepers of N’ (see examples 27–28).
‘O you soul of “The-one-who-has-the-head-of-a-ram”!’
57 For the notion and function of “phrase classifiers,” see footnote 41. 58 E.g., Erman & Grapow (1926–1931, I: 104,3, 164,17). 59 In certain instances, one could alternatively interpret the morpheme spelled n not as the Indirect Genitive
connector n(~) ‘of,’ but as the “dative” preposition n ‘for/to:’ jr(.~)-w o# n=N (the_one_at-M.PL door(M)[SG]
DAT=N), more literally ‘the doorkeeper for/to N’ (still analyzed this way in Werning 2011, II). In example (28),
however, this alternative interpretation is questionable since inanimate dative objects are normally introduced not
by n ‘DAT, for, to’ but by r ‘to, ATTD’ in Earlier Egyptian (Erman & Grapow 1926–1931, II: 193–194, 386–388).
And in (29), this explanation is impossible, since there is no element spelled n at all. 60 Schenkel (2012: ch. 6.3,c); see also Vernus and Polis (both in this volume).
Daniel A. Werning, "Genitive" possessive constructions in Égyptien de tradition: Compound construction vs. of-construction,to be
published in: Eitan Grossman & Stéphane Polis (eds.). Possession in Ancient Egyptian, Berlin: de Gruyter
In an original Earlier Egyptian text, we would have expected the spelling n-ns-sw-w for the
morpheme group n(~)=sw in this possessive construction.61 In this instance, however, the morpheme
sequence is spelled n-s-SCROLL (n(~)=s). The spelling lacks a representation of w and it exhibits an
additional SCROLL classifier. It is obviously derived from the Late Egyptian fossilized morpheme n-
ns-s-~-SCROLL (n(~).sA) ‘belong to; he/she/it belongs to,’ which is the functional successor of Earlier
Egyptian n(~)=sw, n(~)=s~, n(~)=st, as well as simply n(~)= in this construction.62 While the construction in
Caverns is proper Earlier Egyptian, the spelling n(~)=s for the expected morpheme sequence n(~)=sw is
derived from orthography of its Late Egyptian fossilized successor n(~).sA.
6 Conclusions
In the late 2nd and the earlier 1st millennia BCE, authors who wrote texts in Égyptien de tradition, i.e., who
emulated Earlier Egyptian language in their texts, were remarkably well informed about the “ancient”
written language varieties from the 3rd and earlier 2nd millennia BCE. Only constructions that are scarcely
attested and very probably also were scarcely attested in ancient texts also escaped the ancient Egyptian
philologists (Section 3). Nevertheless, the modern linguist can eventually spot influences from the authors’
contemporary chronolect in their Égyptiens de tradition (Sections 2, 4, and 5.3).
In Section 4.2, we explored the usage of the Direct Genitive construction (NPOSSED==NPOSSOR) vs. the
Indirect Genitive construction (NPOSSED n(~) NPOSSOR) in the Book of Caverns (early Égyptien de tradition, 13th
century BCE). It turned out that the authors clearly preferred the Direct Genitive to the Indirect Genitive
(Table 3). Their contemporary chronolect, i.e., Late Egyptian, shows the opposite tendency, i.e., an
increasing preference for the Indirect Genitive. The preference of the Direct Genitive in Caverns is to be
understood as a strategic means to emulate Earlier Egyptian grammar. Moreover, in cases in which the
authors used the Indirect Genitive, they took care to inflect the genitive connector by gender and number
(Table 4), as it is usual in original Earlier Egyptian. This is remarkable since this inflection was not present
in the Indirect Genitive construction of their contemporary chronolect anymore (Table 1).
Furthermore, I highlighted the special effect on statistics of animacy that certain types of genitive
relations have that I tentatively labeled “agentive possesseds” (e.g., ‘ruler of the netherworld’) and
“possessing possesseds” (e.g., ‘possessor of a bier’). In these cases, it is not the constructional possessor but
the constructional possessed which “controls” or “possesses” its nominal partner.
In Section 4.2.1, we explored the variety of more complex instances of genitive constructions, i.e.,
instances in which additional attributes accompany the possessor noun or the possessed noun as well as
instances of nested genitive constructions (‘the N of N’s N’). Most of these match original Earlier Egyptian
patterns. There are, however, some construction variants that are exceptional for — or even alien to —
original Earlier Egyptian (examples 14–19, 29). These are cases for which analyses as true Direct Genitive
constructions, i.e., constructions in which all but one of the nouns are phonetically reduced, are improbable
or even impossible. I suggested that these instances be understood as results from a mechanical translation
of basic contemporary constructions into patterns that the authors attributed to Earlier Egyptian grammar.
We may reconstruct the following rule: “(a) Delete articles; and (b1) inflect genitive connector or (b2)
delete genitive connector.” Note that phonetic considerations are not part of this translation rule.
Consequently, the authors created also constructions like, for example, [N N N] ‘the N of N’s N’ which
61 Gardiner (1957: §114,2). 62 Erman (1933: §233), see also Polis (this volume).
Daniel A. Werning, "Genitive" possessive constructions in Égyptien de tradition: Compound construction vs. of-construction,to be
published in: Eitan Grossman & Stéphane Polis (eds.). Possession in Ancient Egyptian, Berlin: de Gruyter
are largely alien to original Earlier Egyptian since the original Direct Genitive construction entailed the
phonetic reduction of all but one of the participating words ([N\STC==N] ‘the N of N’), which in turn
imposed restrictions on its usage (see Section 4.1). As a result, we may conclude that the authors of Caverns
seem to have conceptualized the original Earlier Egyptian phonetically marked Direct Genitive rather as a
Simple Juxtaposition Genitive Construction, probably not in all but at least in some instances.
Figure 10 maps the translation process as to be reconstructed for the Book of Caverns.
Figure 10. Mental translation of Late Egyptian genitive constructions to the Égyptien de tradition as found
in the Book of Caverns
In Section 4.2.2, we explored possible semantic influences on the choice of either the Indirect or the Direct
Genitive constructions. We found that, contrary to expectations based on iconicity principles, inalienable
possession is expressed comparatively more often with the less compact Indirect Genitive construction (≈
35%) than alienable possession is (≈ 15%), and, accordingly, alienable possession is expressed
comparatively more often with the more compact Direct Genitive construction (≈ 85%) than inalienable
possession is (≈ 65%) (see Figure 6).
Furthermore, I reviewed a mapping of all attested semantic relations in the two genitive constructions in
Caverns onto Nikiforidou’s (1991) semantic map of possessive meanings. To this end, I added the semantic
relation of Entity | Identified (entity) to the map (for some instances, it remains unclear or even difficult to
appoint the sematic relation to a particular class; see Section 4.2.2). In any event, both the Direct Genitive
construction and the Indirect Genitive construction cover large spaces on the semantic map. They are,
however, not identical (compare Figures 8 and 9). Notably, the peripheral meanings Thing constituted |
Constituent material and Entity | Identified (entity) are attested in the Indirect Genitive construction only,
and, as a possessed noun, the kinship term z# ‘son’ is always used with the Direct Genitive construction. As
opposed to typological expectations in the case of natural languages, the semantic space covered by the
Direct Genitive construction does not form one single contiguous space. Besides the main space, there is a
second island. It is unclear whether this is the case due to the fact that the language of Caverns is an
artificial emulation of ancient language varieties or whether the underlying semantic map needs to be
revised.
Specific glossing abbreviations
- affix boundary
= clitic boundary
== Direct Genitive construction
=?= Direct Genitive construction, questionable
Daniel A. Werning, "Genitive" possessive constructions in Égyptien de tradition: Compound construction vs. of-construction,to be
published in: Eitan Grossman & Stéphane Polis (eds.). Possession in Ancient Egyptian, Berlin: de Gruyter
ADJZ adjectivizer (nisbe)
ART article
ATTD attached (adpositional meaning)
COLL collective
DAT dative (function)
DEM demonstrative
DETV determinative article
DIST distal
DISTR distributive
GEN genitive (function)
N noun
NEG negation
PART participle
PTCL particle
POSS possessive
POSSED possessed
POSSOR possessor
PRON personal pronoun
PROX proximate
QUANT quantitative
REL relative pronoun
RELF relative form
SBJV subjunctive
STC status constructus
Abbreviated literature
DZA = Berlin-Brandenburgische Akademie der Wissenschaften. Das digitalisierte Zettelarchiv [des
Wörterbuchs der ägyptischen Sprache] (DZA). In: Thesaurus Linguae Aegyptiae.
http://aaew2.bbaw.de/tla/ (accessed 2014).
L’Égyptologie en 1979 = Unknown editor. 1982. L’Égyptologie en 1979: Axes prioritaires de recherches
(Colloques Internationaux du Centre National de la Recherche Scientifique 595). Paris: Éditions du
Centre National de la Recherche Scientifique.
Hb. + [text number] = Werning (2011: II).
Bibliography
Aikhenvald, Alexandra Y. 2013. Possession and ownership: a cross-linguistic perspective. In Alexandra Y.
Aikhenvald & R. M. W Dixon (eds.). Possession and ownership. A cross-linguistic typology
(Explorations in Linguistic Typology 6), 1–64. Oxford: Oxford University Press.
Allen, James P. 2013. The ancient Egyptian language: An historical study. Cambridge, New York:
Cambridge University Press.
Altenberg, Bengt. 1982. The genitive v. the of-construction: A study of syntactic variation in 17th century
English (Lund Studies in English 62). Lund: CWK Gleerup.
Daniel A. Werning, "Genitive" possessive constructions in Égyptien de tradition: Compound construction vs. of-construction,to be
published in: Eitan Grossman & Stéphane Polis (eds.). Possession in Ancient Egyptian, Berlin: de Gruyter
Berlin-Brandenburgische Akademie der Wissenschaften. Das digitalisierte Zettelarchiv [des Wörterbuchs
der ägyptischen Sprache] (DZA). In: Thesaurus Linguae Aegyptiae. http://aaew2.bbaw.de/tla/
(accessed 2014).
Booij, Geert, Christian Lehmann, Joachim Mugdan, Wolfgang Kesselheim & Stavros Skopeteas (eds.).
2004. Morphologie: Ein internationales Handbuch zur Flexion und Wortbildung. 2. Halbband
(Handbücher zur Sprach- und Kommunikationswissenschaft 17/2). Berlin: Mouton de Gruyter.
Borghouts, Joris F. 2010. Egyptian: An introduction to the writing and language of the middle kingdom
(Egyptologische uitgaven 24). Leiden: Nederlands instituut voor het Nabije Oosten.