"Our language is very literal": Figurative expression in Dene Sųłiné [Athapaskan]

Abstract

For reasons perhaps more sociolinguistic than linguistic, Athapaskan languages rarely borrow in order to expand the lexicon. Instead, they opt for the recycling of a very small set of core stems through recombination or simple reinterpretation (involving metaphor and metonymy). The resulting lexical inventory is striking to cognitive linguists because of the way experiential reality and typologically common construal patterns are routinely exploited for lexical expansion. This paper presents figurative lexicalization strategies in Dene Sųłiné and argues that they are similar to patterns adopted by genetically unrelated languages. Importantly, a deeper understanding of the processes involved in Athapaskan lexicalization may help speakers continue to lexicalize new concepts in indigenous ways, thus helping sustain the health and viability of their languages.

Full text

1
“Our language is very literal”: Figurative expression in Dene Sųłiné [Athapaskan]
Sally Rice
University of Alberta
Abstract
For reasons perhaps more sociolinguistic than linguistic, Athapaskan languages rarely borrow in order
to expand the lexicon. Instead, they opt for the recycling of a very small set of core stems through
recombination or simple reinterpretation (involving metaphor and metonymy). The resulting lexical
inventory is striking to cognitive linguists because of the way experiential reality and typologically
common construal patterns are routinely exploited for lexical expansion. This paper presents figurative
lexicalization strategies in Dene Sųłiné and argues that they are similar to patterns adopted by
genetically unrelated languages. Importantly, a deeper understanding of the processes involved in
Athapaskan lexicalization may help speakers continue to lexicalize new concepts in indigenous ways, thus
helping sustain the health and viability of their languages.
Running Head: Figurative expression in Dene Sųłiné [Athapaskan]
1. Introduction
Like most Athapaskan or Dene languages, Dene Sųłiné resists borrowing as a way of extending
the lexicon.1 A long-discredited hypothesis first advanced by Sapir (1921: 196) holds that
elaborate derivational and inflectional processes within the verbal complex conspire to keep
foreign loan words to a minimum (cf. K. Rice 1989, 2000; Hargus 2007), for descriptions of
some of these processes). However, this morphological resistance may instead be due to socio-
cultural factors rather than linguistic, a hypothesis advanced at the end of this paper. In any
event, language-internal word-formation mechanisms which I will subsume under the rubric
periphrasis (such as relativization, apposition, incorporation, and compounding) as well as
conversion (which can involve special morphology in the way that the plural of computer mouse
in English is often mouses, not mice) seem to be the most productive means of achieving lexical
extension in Dene Sųłiné and its Athapaskan sisters both historically––for indigenous terms––
and synchronically––for terms of acculturation. Illustrative examples of some of these
periphrastic or morphosyntactic pathways of lexicalization are presented in (1):2
1 Literally, ‘the true people’, Dene Sųłiné is the preferred ethnonym for this northern Athapaskan
language loosely associated with the Mackenzie Basin, generally replacing the Cree-based Chipewyan,
except in its ISO 639-3 code: chp. In this paper, I employ the practical orthography used at Cold Lake,
Alberta. That system conflates /e/, /ɛ/, and /ǝ/ and writes them all as e. High tone and nasalization
(represented with a Polish hook) on vowels are phonemic, as are ejectivized versus “plain” stops and
affricates (represented with an apostrophe). Other possibly unfamiliar bigraphs with their IPA values are
as follows: th = /θ/, dh = /ð/, gh = /ɣ/, sh = /ʃ/, zh = /ʒ/, dz = /ʤ/, tł = /tɬ/.
2 I use lexicalization in the sense discussed in Pawley 1985, viz. in a manner that includes both
multimorphemic items such as compounds (i.e., laptop or overindulge) as well as periphrastic items such
as idioms (i.e., nickel and dime to death) which have lexical status.

2
(1) a. nadudh-i ‘the one that slithers along’ ‘snake’ RELATIVIZATION
b. nilts’i slini ‘wind it is evil’ ‘tornado’ APPOSITION
c. sets’é-yałti ‘towards me-3SG talks’ ‘s/he’s scolding me INCORPORATION
d. tthe-sheth ‘stone-hill’ ‘mountain’ COMPOUNDING
e. -la ‘hand’ ‘job’ CONVERSION
There is some evidence that loan translation has also been deployed to achieve lexical
extension especially for many terms of acculturation brought during the first wave of European
colonization, since many languages of native North America use similar imagery and a similar
lexico-semantic “recipe” for the same salient referent (cf. Brown 1999), as shown in (2):
SOME PROBABLE CALQUES
(2) a. kóntué ‘fire water’ ‘alcohol, whiskey’
b. bescho nené ‘big knife country’ ‘America’
c. ghįnaze ‘little worm (pupa, maggot)’ ‘rice’
Nevertheless, there are few true borrowings beyond proper (Christian) names, most of which
derive from French and have been altered to conform to Dene phonotactics. I have encountered
only a few dozen conventionalized loans in over 18 years of interactions with speakers.)
BORROWINGS
(3) a. libada < French le potate ‘potato’
b. masi < French merci ‘thank you’
c. lidi < French le thé ‘tea’
d. Liząbér < French /ɛlízabɛ(t/ ‘Elizabeth’
e. susíkiyás < Cree osikiyâs ‘lizard’
While interesting in their own right, the morphological mechanisms driving lexicalization are
secondary to my purpose here. In this study, I report on lexical extensions and apparent
innovations in Dene Sųłiné that have come about morphologically or periphrastically through the
application of some typologically common metaphors and metonymies. Some preliminary
examples, presented in (4) and (5), are hardly unusual to readers familiar with Lakoff and
Johnson (1980); Panther and Radden (1999); and Panther, Thornburg, and Barcelona (2009). The
actual trope types are expounded on in Sections 2 and 3.
TWO METAPHORS
(4) a. sets’éni ‘the towards me one’ ‘my friend’ IN IS GOOD
b. ets’eze gaiaze ‘little white kidney’ ‘chickadee’ FORM SIMILARITY
TWO METONYMIES
(5) a. bąlai ‘that which is round (‘button’)’ ‘the French’ PART FOR WHOLE
b. nát’adhi ‘that which is cut twice’ ‘square’ PROCESS FOR PRODUCT

3
A case study such as this is intended to demonstrate the ubiquity of figurative processes in
everyday language use while advancing the premise that such processes can be relatively
constrained and systematic in language(s).
The rest of this paper is organized as follows. In §2, I briefly describe some typologically
common metaphors and metonymies that are highly prevalent in Dene languages. I then
summarize some of the typological literature on figurative lexicalization pertinent to the later
discussion on Athapaskan. In §3, I present examples from the Cold Lake variety of Dene Sųłiné,
a fairly conservative but sadly moribund dialect spoken in east central Alberta, Canada. Some
general tendencies are discussed in §4 and compared with examples from other Athapaskan
languages. Finally, §5 addresses issues pertaining to the function and analysis of metaphor and
metonymy in a language’s lexicon and grammar and why it is neither paradoxical nor
oxymoronic for speakers to insist that their very figurative languages are, in fact, very literal.
2. Conceptual and typological patterns of lexicalization
Much early research in cognitive linguistics (henceforth CL) revolves around the study of
grammaticalization and lexicalization patterns cross-linguistically (cf. Talmy 1985, Traugott &
Heine 1991, Heine et al. 1991, Bybee et al. 1994, Svorou 1994). One product of this research has
been to demonstrate how conceptualization and human experience mediate linguistic patterning
(this is the central message in Langacker 1987/1991a, 1991b; Lakoff 1987, Johnson 1987, and
Taylor 1989). Of special interest is exploring how semantic and functional extensions arise for a
given lexical item or construction. Metaphor and metonymy have both been implicated in such
extension processes in language, although they are by no means the only mechanisms of
semantic change. While there are major differences between these processes, they each entail
shifts of reference within or between what philosophers and cognitive scientists call mental
models or what cognitive linguists call background or cognitive domains. That is, the semantic
use or interpretation of metaphors and metonymies involve a projection of language commonly
expressing (usually) more concrete or real-world relations or situations to (often, but not
necessarily) more abstract or idealized cases in which the domain of reference might be ideation,
causation, or textual expression itself. However, metaphors and metonymies are not just used to
describe the abstract or otherwise inexpressible. They can be recruited for purposes beyond the
utilitarian as well; for example, for cultural or metalinguistic reasons, a topic I return to in §5. It
should be pointed out first, though, that inter- and intra-domain projections as implicated in
metaphors and metonymies are responsible for the widespread ambiguity and polysemy found in
language and are a major force in driving semantic change and grammaticalization (cf. notably

4
Lakoff & Johnson 1980; Sweetser 1990, Heine & Kuteva 2002). Many words have multiple
meanings and grammatical functions, although individual meanings are usually specific to a
particular background domain. Critically, the domain of application bears on the intended or
correctly inferable semantic meaning of a term (cf. Croft 1993). This is especially the case with
metaphor and metonymy.
2.1 Metaphor
Literary scholars and cognitive linguists both characterize metaphor more or less the same way:
as an inter-domain mapping function. An expression that has a literal interpretation in one
domain of application takes on a figurative meaning in a second domain. Lakoff and Johnson
(1980) inspired most of the contemporary CL work on metaphor and metonymy (cf. Dirven
1985, Claudi & Heine 1986, Goossens 1990, Langacker 1991b, Radden & Kövecses 1999,
Panther & Radden 1999, Gibbs 2008, and Panther et al. 2009). Collectively, this research not
only has catalogued many conventional metaphors and metonymies across languages, but it has
analyzed them systematically. Lakoff and Johnson placed the study of these figurative tropes
squarely in the realm of linguistic analysis of everyday (rather than specialized or literary)
language. The examples throughout this paper are necessarily of the former type since Dene
Sųłiné has but the shallowest of written traditions and most remaining speakers cannot read or
write using either a practical roman orthography or the Cree-based syllabary familiar to them
from the rather antiquated Roman Catholic (Oblate) hymnals and prayerbooks passed down from
their forebears. Moreover, metaphor and metonymy are ubiquitous in colloquial language and do
not just pertain to a highly composed genre or register. Most examples of everyday metaphors
rarely strike the average speaker as conspicuously figurative or unusual. Indeed, many times the
metaphor has to be stated explicitly before it is recognized as such, a point I return to in §5.
Lakoff and Johnson (1980) identified three highly prevalent classes of conventionalized
metaphors based on correlations we perceive in our experience: orientational, ontological, and
structural (similarity) metaphors. Orientational metaphors obtain when expressions associated
with location or movement along a vertical or horizontal axis signal non-spatial and especially
qualitative relations. Specifically, verbs, adpositions, and adverbials associated semantically with
location or movement upwards or inwards (towards the speaker) in space or along a scale
(always spatially construed) are more positively esteemed than are those associated with location
or movement downwards or outwards (away from the speaker). Ontological metaphors are those
whereby intangible, ephemeral phenomena (like TIME, IDEAS, or EMOTIONS) which frequently
lack direct means of expression in a language, can be talked about and even conceptualized as if
they were substantive, directly perceivable, and imbued with value or other physical qualities

5
like a real concrete object. Structural metaphors fall out under what Lakoff (1987) calls “great
chain of being” metaphors. Associated behaviors or attributes of entities up and down the
epistemological animacy scale can be mapped onto entities in other (both higher and lower)
categories: PEOPLE ↔ ANIMALS ↔ PLANTS ↔ INANIMATE OBJECTS, etc.
2.2 Metonymy
Metonyms are perhaps more ubiquitous in language than are metaphors and individual
metonymies more widespread across languages (cf. Radden & Kövecses 1999, Panther &
Radden (1999), and Panther, Thornburg, & Barcelona (2009)). Metonymy is characterized in CL
as an intra-domain mapping function. Some subpart of a thing or aspect of a relation comes to
stand for the whole in a typical metonymy or, conversely, the whole can stand for a part. Of
special relevance to the Dene Sųłiné examples detailed in §3 are, of course, both PART FOR
WHOLE and WHOLE FOR PART metonymies (of which CONTAINER FOR CONTENTS metonymies
are a special and frequent case), but also GENERIC FOR SPECIFIC and SPECIFIC FOR GENERIC
metonymies, which Sullivan & Sweetser (2010) contend are, in fact, metaphors––a position that
I do not happen to endorse. Perhaps most widespread in the Athapaskan languages are those
metonymies which are verb based, operating on relational predications in one of two ways. In
the most obvious case (since most nouns in the language are deverbal, arrived at morphologically
through the addition of a relativizing or nominalizing particle in a ‘the-one-that-VERBS’ or ‘the-
one-that-is-VERBED’ type of schema), the entire process stands for a salient event participant
bearing an AGENT, EXPERIENCER, INSTRUMENT, PATIENT, LOCATION, or MANNER role, a sub-
part of that process. My corpus is full of such ATTRIBUTE/BEHAVIOR FOR ENTITY metonymies.
Equally robust are PART FOR WHOLE metonymies in which a sub-part of a state or relation or a
sub-phase of some process (such as the initial cause or end result) can come to stand for the state,
relation, or process itself. In most cases, it is an initial phase that stands for the whole. These
metonymies may be less obvious, but they are highly prevalent in Athapaskan languages due to
the ready conversion of noun stems into verb stems and the ubiquity of deverbal nominalizations.
Most of the lexicalizations examined in §3 involve the following patterns of metaphor and
metonymy. Metaphor features prominently in cases of conversion and compounding, in which
some term for a concrete substance may be applied in a more abstract or at least non-literal
domain. Since, as mentioned previously, many nominals in Dene Sųłiné are derived from verbal
sources, PROCESS FOR {RESULT, EFFECT, PRODUCT, AGENT, EXPERIENCER} verb-based
metonymies are extremely numerous. Indeed, verb-based signification (through relativization)
seems much more prevalent than does noun-based (through exogenous compounding or
incorporation) in the language. Some of the diverse lexical domains bearing witness to these

6
figurative lexicalization processes include the very familiar practice of PERSON, GROUP,
ANIMAL, and PLACE NAMING, as well as TOOL or CULTURAL ARTIFACT NAMING. Less
commonly considered cases involve PSYCHOLOGICAL STATES, DISEASES, and various PROCESS
predications. I will exemplify and discuss each of these in turn in §3. First, I present two
proposals about lexicalization patterns cross-linguistically that have tried to account for naming
tendencies in particular.
2.3 Some proposals about lexicalization tendencies
The study of metaphorically and metonymically inspired semantic extension has a huge
literature, not just in the CL world, but in typological and historical linguistics as well. I single
out two sets of investigations in particular because they (a) treat metaphor and metonymy on par
and (b) they are intensely cross-linguistic. Cecil Brown’s (1999) massive study of 77 terms of
acculturation across nearly 200 New World languages dissects naming tendencies by region,
genetic stock, dominant European colonizer, degree of bilingualism, as well as semantic domain
of the artifact. At a coarse-grained level, he divides his concepts into natural kinds,
encompassing introduced fruits (‘watermelon’), vegetables (‘peas’), grains (‘rice’), livestock
(‘pig’) and domestic animals (‘chicken’), and artifacts, such as prepared foods (‘butter’), tools
(‘fork’), storage items (‘bottle’), clothing (‘button’), domestic items (‘candle’, ‘window’),
measurements (‘mile’), and a host of other concrete and abstract concepts (e.g., ‘key’, ‘soldier’,
‘school’, ‘Wednesday’). He reports on percentages of loans, calques, loan blends, or indigenous
lexicalizations inspired by metaphor and metonymy for his many factors and his many items.
Brown subsumes most of his indigenous lexicalization strategies under one of two general types:
referential extension/marking reversal (some type of form-similarity metaphor whereby a native
term is extended, often upon modification, to name the introduced concept as in ‘sun’ for ‘clock’
or ‘big-dog’ for ‘horse’), or descriptive focus/utilitarian naming based on a salient feature of the
item (a kind of PART FOR WHOLE metonymy as in ‘the rounded one’ for ‘button’) or on how
humans use the item (also a kind of metonymy, usually based on a PROCESS FOR PARTICIPANT
metonymy as in ‘that which you write it down with’ for ‘paper’). He concludes there is a robust
correlation between the nature of the introduced items (living thing or artifact) and the nature of
the naming pattern (referential extension/metaphor vs. utilitarian function/metonymy), finding
that only 10% of the items in his corpus that referred to introduced living things were given a
utilitarian name as opposed to 63% of imported artifacts (Brown 1999: 41).
As a simple illustration of the Brown findings, consider the items in Figure 1 which shows
images for two items of acculturation (depending on the culture), one indigenous to native
peoples of North America, although not exclusively so, and the other a modern variant of an

7
artifact first introduced by European colonists. In English, the top item has been lexicalized
through a compound, as shown on the right. Thus, it is multi-morphemic, analyzable, and
figurative by virtue of a GENERIC FOR SPECIFIC metonymy or some other kind of utilitarian
description (‘outer shoe for use with other shoes in deep snow made of bent birch wood and
sinew’), unlike a similar––but borrowed and therefore unanalyzable––monomorphemic English
word, ski. The Dene Sųłiné equivalent, ʔaih, is monomorphemic, unanalyzable and therefore
non-figurative, as befits an indigenous concept to people who traditionally lived in the seasonally
snowy boreal forests of the subarctic. By contrast, either of the English terms for the firearm,
gun or rifle, are arguably unanalyzable and monomorphemic to modern speakers. However, the
Dene Sųłiné equivalent, helk’édhi (< he-l-k’édh-i) ‘that which shoots’, is multi-morphemic and
structured around a PROCESS FOR INSTRUMENT metonymy. In Brown’s (1999) terms, both
snowshoe and helk’édhi would be classified as lexicalizations framed around functional utility,
typical for manufactured cultural objects, as opposed to his “natural kinds.” These contrasting
examples illustrate one purported cross-linguistic lexicalization tendency: that terms for items of
acculturation tend to be borrowed or figurative. In each case, the native object is lexicalized
simply and literally, while the encountered object is lexicalized complexly and figuratively. As
we will see in §3, terms of acculturation are indeed overwhelmingly figurative in Dene Sųłiné,
but strikingly, so are most indigenous concepts as well.
non-figurative lexicalization
figurative lexicalization
ʔaih
[Dene Sųłiné]
snowshoe
lit. ‘shoe for snow’
[English]
gun, rifle
[English]
helk’édhi (t’elk’idhi)
lit. ‘the thing that shoots’
[Dene Sųłiné]
Figure 1. Figurativity in English and Dene Sųłiné lexicalization for two terms of acculturation.
In a similar vein and inspired by some of Brown’s earlier work with his colleague, Stan
Witkowski (Witkowski & Brown 1978), David Wilkins (1996) searches for cross-linguistic
tendencies of semantic change (what we could also call semantic extension leading to polysemy)
in the admittedly circumscribed but completely universal and therefore always indigenous
domain of “parts of a person”. Tracking 75 concepts across a large number of languages of

8
central and western Australia, he concludes that metaphor and metonymy are differentially
responsible for the most commonly attested semantic extensions or chains (e.g. ‘egg’ →
‘testicle’ vs. ‘skin’ → ‘body’ → ‘person’) affecting body part naming. Although his larger aim is
to better understand semantic shift in order to expand the list of possible cognates for purposes of
proto-reconstruction across a language family, he does propose an implicational hierarchy within
the semantic field of “parts of a person” which could be tested both cross-linguistically and
across other semantic domains. This hierarchy, presented in Figure 2, is especially relevant in the
context of the present volume since it puts metonymy alongside metaphor as a patterned and
cognitively motivated mechanism of semantic change in language after language.
intrafield
metonymic changes
→
interfield
metonymic changes
→
interfield
metaphoric changes
→
intrafield
metaphoric changes
‘skin’ → ‘body’
‘smell’ → ‘nose’
‘spear’ → ‘penis’
‘anus’ → ‘mouth’
Figure 2. Wilkins’ (1996: 274) four classes of semantic change, ranked hierarchically, within the
semantic field “parts of a person.”
The present study of figurative lexicalizations in Dene Sųłiné ranges beyond the typical
referential denotata of the Brown and Wilkins investigations. Although I do include terms of
acculturation and body parts, I also investigate figurativity in indigenous concepts as well as in a
host of relational predications (states and processes) in the language.
3. A Semi-Structured Inventory of Metaphors and Metonymies in Dene Sųłiné
In this extended section, I survey certain semantic domains in Dene Sųłiné that are replete with
lexicalizations based on metaphor, metonymy, or a combination thereof. This is by no means an
exhaustive inventory––one feels as if the surface has barely been scratched––but I believe it to
be representative. Moreover, I make no claims about the figurative uniqueness or universality of
these expressions. That is, the lexical formatives and/or conceptual imagery contained in the
following expressions may or may not be particular to this dialect or this language. The resulting
lexicalization may involve nothing more than a wholesale calque or loan translation from other
languages. Nevertheless, an ingenious combination of indigenous lexical items driven by
metaphor and/or metonymy has produced an impressive set of innovative and often idiomatic
expressions. Taken together, they contribute to a line of argumentation in CL that seeks to
demonstrate the central role that meaning plays in lexical and grammatical structure in language.
They also bring us closer to understanding the cognitive means by which human beings, no
matter the culture, come to linguistic terms with the world around them and within them.

9
In the case of Dene languages, the stem inventory is staggeringly small. Estimates range
from 1300-2000 semantically discernible (though often very vague), relatively cognate,
phonologically coherent, and generally monosyllabic lexical stems (Victor Golla and Jim Kari,
p.c.). Semantic extension (leading at times to cross-categorial conversion with attendant
morphological adjustments) appears to be fairly robust, but it remains a poorly studied part of the
Athapaskan lexicon. It is especially the case that verb stems, often highly suppletive, will “cross-
lexicalize”.3 There may be arguments for treating these stems as highly polysemous or at least as
engendering chained associations via metaphor and metonymy. I tackle Athapaskan verb stem
polysemy at length in Rice (forthcoming). Nevertheless, Dene verb stems––and not just the well-
known classificatory verb stem system––are notoriously vague and generic, and only gain their
specificity through a variety of prefixes or context of use.4
Because of the small inventory, stems are routinely called upon semantically to do double
and triple duty, if not more, through conversion, compounding, juxtaposition, and inflection. The
small inventory extends as well to a small set of items that would traditionally be considered
derivational material, encompassing things like augmentatives, diminutives, defunctives, gender
markers, intensifiers, negativizers, nominalizers, qualifiers, and the like. These items, all
suffixes, are still highly productive, but they have allowed for, either singly or in combination,
the creation of many entrenched and conventionalized lexical items, which in turn can be
examined for their degree of figurativity since many exhibit striking metaphors and metonymies.
It is the inventory and analysis of this relatively small set of items that we delve into here.
A word first about format. In individual examples, a metaphor will be identified using an [X
IS Y] comment, while a metonymy will be specified by the rubric [X FOR Y]. If the English (free)
translation is itself figurative, double asterisks (**) will follow the gloss. To save space, I present
data in columnar format, with the Dene example listed first, the literal gloss in the middle
column, and the figurative or free gloss at the right. I primarily present examples of Dene Sųłiné
figurative lexicalizations by semantic domain, regardless of whether they involve pure metaphors
or metonymies. In §4, however, I summarize with comments about the most systematic
metaphors and metonymies observed in the data.
3 For example, the stem -da shows up as the singular imperfective form for ‘sit’, ‘go’, and ‘rock (back and
forth)’. Likewise, the stem -ʔį is associated with paradigms for ‘see’, ‘look around/for’, ‘notice’, ‘wait
for’, ‘steal’, and ‘hide’ (cf. Li 1933).
4 The classificatory verb system conflates position, dislocation, and controlled handling of objects which
are variously construed as stick-like, flat and flexible, solid and round, granular, animate, contained in an
open container, contained in a closed container, and so on. The different Dene languages feature different
inventories of classified objects. The Athapaskan literature features many studies, although Rice
(forthcoming) argues against its exclusivity. I contend that the majority of verbs are classificatory in that
they conflate information about a salient event theme and the event or relation itself.

10
3.1 Naming others and describing the human condition
A lexical domain especially rich in metaphors and metonymies involves (proper) naming. This is
a good category with which to begin because names are both highly conventionalized and highly
charged in terms of cultural identity. Having an epithetic quality as they do, ethnonyms
especially can serve to identify both the referent and the labeler as members of a specific group.
The practice of giving descriptive (that is, figurative) sobriquets or nicknames is typical of
Athapaskan, not to mention Amerindian languages generally (cf., notably, Sapir 1923, 1924;
Young and Morgan 1987: 811-812b; Basso 1990). I will concentrate here on ethnic and group
naming. Proper names, peoples (tribes/nationalities/ethnic groups), and place names are rarely
monolexical. Generally, the resulting composite lexicalization describes something about the
people themselves, a geographical feature of where they come from, or activities or artifacts
native to the region. Most of the ethnonyms in my Dene Sųłiné corpus involve metonymies,
typically either a GENERIC FOR SPECIFIC or a PART FOR WHOLE metonymy.
3.1.1 Ethnonyms
Across the larger Dene world (at the time of European contact, it stretched from Alaska to
Mexico in latitude and from the Pacific Ocean to the Hudson Bay in longitude), the word for
person or people is highly cognate: diné (Navajo), -t’ina (Tsuut’ina), denae (Ahtna). According
to Victor Golla (p.c.), the etymology of the probable proto-form strongly suggests a derivation
based on a stative predication––an ideophone, really––of the form ‘sounds like X’:
(6) dene (<de-na)
lit. ‘the one who sounds human’
fig. ‘Dene person (the one that speaks like a human being)’
This etymological hypothesis is intriguing as it both makes morphological sense and conforms to
what I have found to be a quite common set of metonymies across the many Dene languages that
I have examined: SPECIFIC FOR GENERIC or GENERIC FOR SPECIFIC metonymies. These tropes
are especially common in ethnonymic naming. In the examples below, I gloss the Dene Sųłiné
exponents dene and -t’iné as ‘person’ or just list the group modifier. As singular and plural are
not specially marked in Dene Sųłiné, these terms also refer to the entire ‘people’ so designated.
GENERIC FOR SPECIFIC
(7) a. dene sųłiné ‘genuine/true person’ ‘Dene Sųłiné’
b. ˀena ‘enemy’ ‘Cree’; ‘non-Dene native’
c. hotélna [< hotéli ˀena] ‘barrens (AREAL cover) enemy’ ‘Inuit’
d. des ną(t’iné ‘river-across-people’ ‘Slavey’
e. ˀasi dene ‘some person’ ‘non-native person’

11
With the first (fur traders) and second (homesteaders) waves of colonization, came ethnonyms
based on some salient feature of the interlopers. Usually these features pertained to dress or
lifestyle, hence, they are metonymic. I make no claims about the semantic uniqueness of these
lexicalizations. It is highly likely that most of the examples in (8)-(9) are calques.
ASSOCIATED ARTIFACT (PART) FOR OWNER/WEARER (WHOLE) FOR SPECIFIC
(8) a. tthot’iné [< tthé-yoh t’įné] ‘stone-house person’ ‘English’
b. bescho dene/t’iné ‘big knife person’ ‘American’
ATTRIBUTE FOR WHOLE
(9) a. betthighe tł’ųle nenedhí ‘those whose tied hair is long’ ‘Chinese’
b. ʔįłt’eri nade ‘the ones who are naked’ ‘Ukrainians’
3.1.2 Kith and kindred
Dene people were traditionally hunter-gatherers who migrated seasonally in small multi-family
bands. Kin systems were somewhat fluid (Ives 1990) and the nomenclature system is complex,
although not particularly figurative. Contemporary Dene Sųłiné speakers refer to their family and
fellow band members in similar ways, no doubt since both groups were traditionally their
relatives. Both terms are metaphoric and suggest unity. As shown in (10), one invokes a body
part metaphor whereby the family is construed like a hand (an intriguing source image schema
since its individual parts are as salient as the whole); the other stresses the oneness of the group
in a kind of DIVERSITY IS UNITY image.
(10) a. selot’iné [< sela-hot’iné] ‘my hand/partner-people’ ‘my relatives’
b. įłá dene ‘one people’ ‘family’
Other significant relationships that are lexicalized figuratively in Dene Sųłiné involve
forebearers and descendents, which likewise can be metaphorically (and spatially) construed in
English. In Dene Sųłiné, both orientation and botanical metaphors are at play:
TIME IS SPACE (PAST IS AHEAD or PAST IS BEHIND)
(11) a. tthéridene ‘first people’ ‘ancestors/forebearers’
b. yanísot’ine ‘the past/long ago people’ ‘ancestors’
c. ʔąłnetthi ‘the one who went the length’ ‘elder’
PEOPLE ARE PLANTS
(12) a. betthúe ‘3SG-branch’ ‘his/her grandchildren’
b. bechįghaé ‘3SG-wood-root’ ‘his/her descendants’
With respect to significant relationships of the same generation, a few are lexicalized
figuratively, notably, the concepts of ‘partner’ and ‘friend’. These lexicalizations are based on

12
two metaphors seen previously: a body part metaphor (USEFUL PERSON IS USEFUL BODY PART)
and a spatial metaphor (IN IS GOOD), as shown in (13) and (14), respectively:
(13) sela ‘my-hand’ ‘my same-sex cousin, helper/partner’ (♂ speaker)
(14) sets’éni ‘the-towards-me-one’ ‘my friend’
Although there is no gender differentiation in the personal and possessive pronoun system for
third person singular in any Athapaskan language, there is ample differentiation when referring
to males and females. The generic ‘human/person/people’ term, dene, does not extend
exclusively to males. Although the etymology of the suffix in the male human term, deneyu, is
opaque, it might be related to sį-yeze, ‘my son’ (lit. ‘my little man’?) and -yane, the all-purpose
male suffix used for animals. By contrast, the female term, shown in (15), is strongly metonymic,
based on a PART FOR WHOLE metonymy. The -kwi suffix is an obscure pluralizing morpheme
which applies to human collectives (like the English -folk), although it and its cognates in other
Dene languages are no longer very productive.
(15) ts’ékwi (< ts’ér-kwi) ‘womb-ones’ ‘woman, women’
Athapaskan kin systems are fairly complex, with differences in cross-ness and parallel-ness
(sic) extending across three generations, as well as differentiation in older and younger siblings.
Thus, while there are many distinct terms, there is also some morphological recycling, as the
examples in (16) show. These could probably be considered examples of what Brown (1999)
calls a marking reversal or what Wilkins (1996) calls an intra-domain metaphor. Within the
content domain of kinship, the diminutive singles out individuals who are of different
generations than those referred to by the non-derived stems, but not necessarily descending
generations. Nevertheless, the derived forms do suggest an especially close relationship to ego,
as is often the case when a diminutive is used. This lexical extension via the diminutive gives
rise to what I will call a SMALL IS FAMILIAR metaphor in the context of kin terms.
SMALL IS FAMILIAR
(16) a. sunaghaze [<sunaghe-aze] ‘my little older brother’ ‘my grandson’
b. setáze [<setá-aze] ‘my little father’ ‘my uncle (father’s brother)’
c. sáraze [<sáre-aze] ‘my little older sister’ ‘my granddaughter’; also ‘my
daughter-in-law’ (♀ speaker)
d. setsǫaze [<setsų-aze] ‘my little aunt (dad’s sister)’, ‘my sweetheart’ (♂ speaker)
‘my little mother/sister-in-law’,
‘my cross-sex cousin’ (♂ speaker)

13
3.1.3 Cultural roles
Pre-contact Dene society was largely egalitarian, with any differentiation reserved for chiefs and
shamans (Abel 2005). Both of these traditional societal roles are lexicalized via a relativization
(‘the one who Vs’); thus, they are based on a deverbalized process and are therefore metonymic.
Just as a host of non-traditional concrete objects that were introduced into Dene culture required
lexicalization, so too did non-traditional social roles, job titles, or professions. Many of these are
based on the ‘chief’ formative derived from an all-purpose verb of being/doing/acting upon
(which we will revisit later), -dher/-dhi: k’ódheri [< k’á/k’é hólderi], lit. ‘the one who acts
for/on (unspecified)’, fig. ‘chief, boss, ruler, Lord’, as shown in (17). Others involve compounds
with dene ‘person’ or other types of relativizations based on processes (being at, knowing,
speaking, teaching, making, etc.).
GENERIC FOR SPECIFIC or SPECIFIC FOR GENERIC
(17) a. k’ódheri ‘chief’ ‘factor (head of fort or trading post)’
b. tsąba k’ódheri ‘money chief’ ‘Indian agent (dispenses treaty money)’
c. dení k’ódheri ‘moose chief’ ‘forest ranger/game warden’
d. k’ódheri nethé ‘chief important’ ‘king/prime minister’
e. k’ódheri nethé ts’ékwi ‘chief important woman’ ‘queen’
PROCESS FOR AGENT and GENERIC FOR SPECIFIC
(18) a. yałti ‘the one who speaks’ ‘priest’
b. tthethiyį ‘the one who stands (at the) head’ ‘leader’
c. tsątsáné k’olyąi ‘the one who knows about metal’ ‘mechanic’
d. níhołtsįni ‘the one who made earth’ ‘the Creator’
DOING IS BEING AT and BEING AT IS BEING
(19) a. *nádher ‘3SG stays/lives customarily’ [infelicitous without a complement]
b. łueghąnádheri dene ‘the person who’s about fishing’ ‘fisherman’
c. įłts’uzi gáh nádher ‘3SG lives about the trap’ ‘trapper’
d. seʔą( nádheri ‘the one who stays by me’ ‘my neighbor’
e. įk’ązį( nádheri ‘the one who stays (about) spirit’ ‘medicine man/shaman’
3.2 Body parts, functions, and dysfunctions
Owing perhaps to the chronically small stem inventory, there is a general lack of differentiation
between humans and animals in Dene Sųłiné when it comes to partonymic naming of the body,
to the point that it is difficult to state whether the terms are simply very vague or, if not, whether
the semantic extension maps from humans to animals or vice versa. While animal and human
body part terms are largely shared, there is substantial differentiation in the verbal lexicon when
it comes to the “selectional restrictions” holding between verbs of position, motion, and even

14
consumption and the location of their subjects on an animacy hierarchy.5 As a case in point,
consider the examples in (20), each of which subsumes a part (or effluvium) that could belong to
a variety of fauna, from insects, fish, and reptiles to birds, animals, and humans (as enumerated
in parentheses).6 The possessive prefix e- in all the examples in (20) is non-specific and
generally used for non-human possessors. The series se-/sį-, ne/nį-, and be-/bį- constitute the
personal possessive prefixes for 1SG, 2SG, and 3SG, respectively.
SPECIFIC FOR GENERIC
(20) a. edheth ‘its-skin/hide/pelt’ (animal, human)
b. edé ‘its-horn/antlers/feelers/antenna’ (animal, insect)
c. eché ‘its-tail’ (animal, bird, fish7)
d. egha ‘its-fur/hair’ (animal, human)
e. ekegané ‘its-claw/hoof/talon/toenail’ (animal, bird, human)’
f. etsáné ‘its-excrement’ (animal, bird, insect, human)
3.2.1 Body part naming
Despite the vagueness alluded to above resulting from overlap between humans and non-humans
in body part naming, there is a lot of figurativity in Dene Sųłiné body part terminology
(examined here) as well as certain bodily secretions and wastes (described in §3.2.2). At play,
largely, is form similarity—with plants (21), animals (22), and other body parts (23).
5 For example, heya ‘3SG goes/sets out’ could be used to describe the motion over land of either a single
human or a single moose (hence two-legged and four-legged motion are conflated here), but a horse or a
caribou or buffalo would require different motion verb stems. As per fn. 4, most position and dislocation
(motion) verbs are classificatory and they reference diverse shape, constituency, and animacy distinctions
in the stems themselves. It should thus not be terribly surprising that there is more specificity in the Dene
verb stem than in the noun stem, as evident in these generic body part terms. See also fn. 15.
6 In perhaps the most striking case I have encountered of semantic chaining or lexical extension across
body parts of different species of different phyla of the animal kingdom, the same body part term, s-tłólé
or s-tłure, which I will gloss loosely as ‘lobe’, shows up in the Dene Sųłiné expressions for ‘(human) ear
lobe’, ‘chicken crop/croup/craw’, ‘turkey’, and ‘(seal) flipper’: as in dene dzastłólé ‘human ear-lobe’,
k’ásba dastłólé ‘chicken chin-lobe’, dastłúré ‘chin-lobe’ (in a complete PART-FOR-WHOLE metonymy
whereby the most salient part of the bird stands for the bird itself), and lastłúré ‘hand-lobe’, as in a seal’s
flipper. Despite minor phonological variations in the shape of this stem morpheme, each instance features
an -s- pre-stem interfix, whose presence between the possessor body part (dza-, da-, and la-) and the
profiled subpart (-tłólé or -tłure) is morphologically licit, but highly marked semantically. I have rare
examples in my field notes of sįstsi instead of the expected sįtsi ‘my nose’ or besdlok instead of the
expected bedlok ‘his/her laugh’ accompanied by comments from consultants saying that the s-marked
body part was ‘cute’, ‘deformed’, or ‘distinctive’ in some way. The late Bob Young (p.c.) called it the
“spurious s”, as he had encountered it in Navajo, as well, where it had much the same function.
7 Northern Dene people are riverine people and fish is a dietary staple. It is not surprising, then, that many
fish part terms are monomorphemic and/or specific to fish: łue gaye ‘fish fin’, łue k’ésé ‘fish gill’, łue
gų(thé ‘fish scale’, and łue k’oné ‘fish eggs’ (although the latter might derive from a form-based metaphor
and literally mean ‘fish fire’, since roe are generally red).

15
PEOPLE ARE PLANTS
(21) a. sįchéné ‘my-wood/stick/branch’ ‘my arm’
b. sekechéné ‘my-foot-wood’ ‘my ankle’
c. betthú ‘3SG-(wood) knot’ ‘his/her/its tongue’
d. benąjulé [also: dene jul] ‘3SG-pinecone’ ‘his penis’
e. bedagheztł’ok ‘3SG-mouth-around-grass’ ‘his beard/moustache’
PEOPLE ARE ANIMALS
(22) a. setthíghá ‘my-head-fur’ ‘my hair’
b. begheze ‘3SG-egg(s)’ ‘his testicle(s)’
PARTS ARE OTHER PARTS (intra-domain metaphorical mapping based on form similarity)
(23) a. setth’utthila ‘my-breast-head-hand(extremity)’ ‘my nipple’
b. betth’ene ‘3SG-bone/skeleton’ ‘his/her leg’
c. setth’itth’ene ‘my-head-bone’ ‘my skull’
d. sets’atth’éné ‘my-hat-bone’ ‘my forehead’
The examples in (24) and (25)-(26) below are all cases of intra-domain mappings described at
length in Wilkins (1996). Although he posits an implicational hierarchy indicating a universal
directionality for certain part-to-part mappings (upper body to lower body, as in WRIST → ANKLE
or FINGER → TOE, or visible to invisible, as in PALM → SOLE), we can not really establish with
any certainty the directionality of a few body part extensions in Dene Sųłiné. For example, -tł’á,
is a morpheme that figures in lexicalizations for ‘buttocks’, ‘cheek’, ‘palm’, and ‘sole’. These
items are commonly construed and lexicalized similarly cross-linguistically because their relative
flatness and fleshiness. Dene Sųłiné is no exception. However, it is difficult to determine
whether the morpheme, -tł’á, has a single basic meaning from which the others have extended,
especially since dene tł’átthén, lit. ‘person’s tł’á-flesh’, can mean either ‘cheek’ or ‘buttocks’. I
would submit that there are more simple compounds with -tł’á that suggest it is the latter (which
goes against Wilkins’ upper-to-lower hypothesis), as shown in (24e-h). Nevertheless, the
morpheme also refers to ‘palm’ and ‘sole’ in the proper contexts.
FORM SIMILARITY
(24) a. -tł’á ‘butt/cheek’
b. dene tł’átthén ‘person’s butt/cheek-flesh’ ‘buttocks’, ‘cheek’
c. denį tł’ághé ‘(in) person’s (hand) butt/cheek’ ‘palm’
d. dene ketł’á ‘person’s foot butt/cheek’ ‘sole’
e. ts’i tł’ághe ‘boat’s butt’ ‘keel’
f. tł’áʔih ‘butt-garment’ ‘pants, trousers’
g. sekwi tł’ádhéth ‘child’s butt-cover’ ‘diaper’
h. tł’áreghesdá ‘butt-I sat’ ‘I rested’
i. betł’á yeghiką ‘I put it in 3SG hands’ ‘I handed it to him/her’

16
We’ve seen from the example in (24b) above that -tthén means ‘flesh’ generally. When modified
by another body part, it can specify particular fleshy regions of the body, as in (25):
(25) dene ghú tthén ‘person’s tooth flesh’ ‘gum (of mouth)’
I would also consider the semantic extension in (26a) as a case of intra-domain metonymic
mapping, although the extension in (26b) does involve a shift from the purely physical to the
more abstract inter-personal or social sphere and, thus, suggests a metaphorical extension from
body to self as well. We frequently tap or place our hand on our chest when introduced to
strangers as we speak our name. In Dene Sųłiné, this semantic shift and relexicalization is
reinforced by the use of the (less-frequent) nasal allomorph of the possessive prefix.8
PART FOR PHYSICAL WHOLE (26a) → PHYSICAL WHOLE FOR ABSTRACT WHOLE (SELF IS BODY) (26b)
(26) a. sezí ‘my-chest’ ‘my body (living)’
b. sįzí alternate possessed form of ‘my-chest/body’ ‘my name’
Other body part terms involve a combination of metaphors and metonymies or, at the least, a
combination of metonymies. In (27), we find a PROCESS FOR PRODUCT metonymy designating
the nail as the dried part of a digit; however, no digit is explicitly mentioned. Instead, the WHOLE
(hand/foot) stands for the PART (finger/toe) and a GENERIC expression, -gané ‘dried’, that’s also
used for dried fish, meat, berries, or even paint, designates the nail––something SPECIFIC. Not
surprisingly, as indicated in (20) above, many body parts extend to both humans and other living
creatures. Toenails are no exception and they also designate claws, hooves, and talons, especially
when used with the non-specific 3SG possessive pronoun e-.
PROCESS FOR PRODUCT, WHOLE FOR PART, and GENERIC FOR SPECIFIC
(27) a. belagané ‘3SG-hand-dried’ ‘his/her fingernail’
b. bekegané ‘3SG-foot-dried’ ‘his/her toenail’
c. ekegané ‘its-foot-dried’ ‘its claw, hoof, talon’
In (28), I list some other figurative body part terms next to a brief descriptive label indicating the
metaphor or metonymy involved.
(28) a. benałchethé ‘3SG’s sack’ ‘his genitals’ FORM SIMILARITY
b. setthíghą( ‘about my head’ ‘my brain’ CONTAINER FOR CONTENTS
c. sedzagór ‘my lower leg-spear’ ‘my knee’ ARTIFACTUAL FOR NATURAL
d. seberts’ene ‘that which is round ‘my navel’ PROCESS FOR THEME
[<seber ts’į ʔane] from my belly’
8 Note the two allomorphs for the possessive prefix for 1SG––a nasal version, sį-, in (26b) and (29d), and
an oral version, se-, in (26a) and (29e). Dene languages often exploit allomorphy such as this for
(re)lexicalization purposes.

17
Certain body part terms (e.g. ‘hand’, ‘foot’, ‘skin’, ‘guts’, and the previously discussed
‘buttocks’) enter into a number of expressions that either make reference via form similarity to a
non-body part (something concrete as in an associated artifact or something a bit more abstract
as in a remnant such as a print or track) or they enter into predications for processes in which the
body part is somehow salient (e.g., expressions of position, transfer, ideation, emotion). Most of
these are metaphorical as well as metonymic. Because they are so numerous and diverse, I offer
no further analysis beyond their strategic listing and glossing.
(29) a. -la ‘hand’ ‘work’, ‘end/extremity’
b. denelachédh ‘person’s hand-duck’ ‘thumb’
c. denelatthałé ‘person’s hand-awl’ ‘finger’
d. sįla ‘my hand’
e. sela ‘my job/work/partner’
f. tłolá ‘grass-end’ ‘grain, wheat’
g. yohtthílá ‘house-head-end’ ‘roof (peaked)’
(30) a. -ké ‘foot’, ‘paw’ ‘shoe’, ‘track’
b. dene kéłts’élé ‘person’s foot-DIM’ ‘person’s toe’
c. dene ké ‘person’s foot’ ‘person’s shoe’
d. ké sųłiné ‘genuine/real shoe’ ‘moccasin’
e. kéchogh ‘shoe-AUG’ ‘boot’
f. dene láké ‘human hand-foot’ ‘finger print’
g. łįké ‘dog paw’ ‘dog tracks’
h. náské ha ‘I set out footing (it)’ ‘I will track (him/her/it)’
(31) a. -chane ‘guts, intestines’
b. echǫtth’éné ‘its-gut-bone’ ‘ribs’
c. echą(theda ‘3SG (baby) is gut-sitting’ ‘she is pregnant’
d. echánestį ‘I’m gut-lying’ ‘I’m lying on my stomach’
(32) a. -dhéth ‘skin, hide, pelt’ ‘cover/sheath’
b. tsádhéth ‘beaver pelt’ ‘fur’ (generic commodity)
c. nadhéth ‘eye cover’ ‘eyelid’
d. dene dadhéth ‘person’s mouth cover’ ‘lip’
e. dene tthí dhéth ‘person’s head cover’ ‘scalp’
f. lezdhéth ‘urine cover’ ‘bladder’
g. jíegaié dhéth ‘bean/pea cover’ ‘pod’
h. k’aigúé dhéth ‘caterpillar cover’ ‘cocoon’
i. t’elk’ithi dhéth ‘gun cover’ ‘holster’
j. tsąba dhéth ‘money cover’ ‘wallet’
Finally, in (33), I list some assorted examples of material objects, either man-made or found in
nature, like those in (32), which are constructed from a body part term which serves as a
metaphorical source for some artifact part.

18
(33) a. yodą(báli ‘house-mouth-canvas’ ‘door’
b. tthįdá ‘headway-mouth’ ‘doorway’
c. įdzíaze ‘its little heart’ ‘strawberry’
d. ebą(dzaghéjeré ‘its rotten rounded ear’ ‘mushroom’
e. ts’ąk’áni k’oth ‘metal-burning (stove) neck’ ‘chimney, stovepipe’
3.2.2 Effluvia
Most effluvium terms in Dene Sųłiné are monomorphemic and, thus, unanalyzable. While both
lez ‘urine’ and del ‘blood’, each a quintessential effluvium, are nouns cognate with verb stems,
neither can be possessed as nouns (the same goes for kuł ‘vomit’), although both stems do enter
into figurative compounds, as in lezdhéth (lit. ‘urine-skin/sheath’, fig. ‘bladder’), or process
metonymies, as in del níłtth’éli (lit. ‘blood that has flowed together’, fig. ‘blood clot’). In any
case, with verbal –lez, as in (34), we have an interesting chicken-and-egg conundrum: Is the
proper basic gloss ‘drip’ or ‘urinate’? If the former, then the expressions in (34b-c) seem very
literal (‘rain drips’, ‘blood drips’), while (34d) is merely vague in a GENERIC FOR SPECIFIC
metonymic way (‘s/he’s dripping’). On the other hand, if the prototypical gloss of -lez has to do
with ‘urinate’, then (34b-c) are quaintly metaphoric in a GREAT-CHAIN-OF-BEING sort of way
(the environment is animate) or in an intra-domain (BLOOD IS URINE) and inter-domain (RAIN IS
URINE) metaphoric way. The expressions in (34a) and (d) do contrast morphologically (and
hence, semantically) on the basis of the pre-stem syllable; de- suggests an impersonal and
intransitive process, while he- suggests a human actor and an imperfective process which might
be transitive or intransitive. This contrast should reinforce a message to both linguists and
learners that most lexical items in the language are constructions and therefore need to be
analyzed holistically and in context, in spite of (or perhaps because of) the morphological
overlap of certain stems.
(34) a. delez ‘it is dripping’
b. chą delez ‘it’s dripping rain’ ‘it’s raining hard’
c. del delez ‘blood is dripping’ ‘s/he’s bleeding’
d. helez ‘3SG is urinating’
A variety of other effluvium terms either have a figurative origin or they give rise to other
figurative expressions. The term for ‘breast milk’, shown in (35), is simply the independent form
of ‘breast’ (for inalienably possessed items like body parts, the independent form features no -é
suffix). The examples in (36) and (37) present some other body-based effluvium extensions.
CONTAINER FOR CONTENTS
(35) tth’u ‘unbound form of breast’ ‘breast milk’

19
EFFLUVIA ARE OTHER EFFLUVIA (intra-domain metaphorical mapping based on form similarity)
(36) a. edzaghé tsané ‘ear excrement’ ‘earwax’**
b. sedayíhé ‘my mouth-breath’ ‘my voice’
EFFLUVIA ARE WATER (inter-domain metaphorical mapping based on form similarity)
(37) a. senatúé ‘my eye-water’ ‘my tears’
b. hetsąnetué ‘defecation water’ ‘diarrhea’
On analogy with (37a), I present (38a) along with a host of other expressions in (38b-m) built on
the stem -jer, which enjoys a considerable vagueness or ambiguity among speakers, no doubt on
account of the presence of this very salient and morphologically stable syllable in a wide range
of expressions. Nevertheless, they all seem to share a certain semantic essence of being rotten or
repulsive as well as being feared. I would argue here that, whatever the core meaning might
indeed be, one sense is clearly ‘fart’, arguably a type of effluvium. I suggest that the expressions
in (38) are probably related through a variety of metaphors and/or metonymies, although I have
no basis for positing any directionality.9
(38) a. najér ‘eye-fart/crud’ ‘sleep/sand’**
b. hesjer ‘I farted’
c. nesjer ‘I’m starting to fart’ ‘I’m afraid’
d. bech’á nesjer ‘I’m afraid against him/her’ ‘I’m afraid of him/her’
e. hojere ‘it’s dirty, messy, rotten’
f. hónejer ‘it’s dangerous’
g. dełjere ‘it smells rotten’
h. ejere ‘that which farts/is rotten, stinky’ ‘cow’
i. bek’eghįłjer ‘it’s rotting on 3SG’ ‘his/her flesh is rotting/scabby’
j. dene bek’eghįłjeri ‘the person who it’s rotting on’ ‘leper’
k. k’eljeri ‘that which is rotting on (s.o.)’ ‘leprosy’
Finally, the term for ‘excrement’ tsą( (or -tsáné in its combined form) gives rise to at least two
non-body-based extensions for food terms, presumably based on a form similarity metaphor. I
hesitate to include the generic word for ‘metal’ here, tsątsáné, although it conforms to the same
pattern as the pair in (39a-b).10 The remaining examples in (39c-f) also involve effluvia-based
metaphors for introduced foodstuffs, based on blood, milk, and rendered fat (grease).
9 A ‘rotten’/‘be afraid of’ polysemy holds in a host of Athapaskan languages. For example, of the 15
languages Hoijer 1956 examined for degree of cognation across a 100-item Swadesh-esque list, 12
featured the same stem for both, data were missing for the ‘rotten’ term in 2 of the languages, while only
1 had a different stem for the ‘rotten’ and ‘be afraid of’ terms. In all 15 languages, however, the ‘be afraid
of’ stem was cognate with Dene Sųłiné -jer.
10 The modifying morpheme tsą (no high tone) is not interpretable as a noun, although low-tone -tsą is
the verb stem for ‘defecate’. Whatever it means, it also shows up in the word for ‘money’ tsąba, which
seems to have the shape ‘X-for’. Could the word for metal, tsątsáné, be reduplicative, as in ‘shit shit’?

20
FORM SIMILARITY (FOOD IS EFFLUVIA)
(39) a. tł’izitthoétsáné ‘bee-excrement’ ‘honey’
b. k’asbatsáné ‘chicken-excrement’ ‘mustard’
c. ejeredelé ‘cow-blood’ ‘chocolate’
d. ejeretth’úaze ‘little-cow-milk’ ‘canned milk’
e. ejeretth’úétłesé ‘cow-milk-grease’ ‘butter’
f. ejeretth’úé níłtth’éli ‘cow-milk that flows together’ ‘(cheese) curds’ [cf. (44f)]
3.2.3 Ailments and diseases
Generally, Dene Sųłiné does not deploy a POSSESSION metaphor the way English does for
expressing illness or some physiological condition. In English, one generically has measles,
cancer, a stroke, a heart attack, an infection, an operation, a skin rash, an incurable disease, a
cold, a headache, etc. Except for a single example in my corpus (in 40b), possession does not
enter into the naming or predication of disease. Moreover, possession in most Dene languages is
handled via a spatial metaphor, specifically a ‘from’ or ablative image schema (cf. Rice 2004), as
shown in (40). The disease expression in (40b) is also based on a CAUSE FOR EFFECT (or at least
a PART FOR WHOLE) metonymy, with suga ‘sugar’ standing for the disease, presumably because
of the well-known role that diet and blood chemistry play in the disease.
HAVING IS BEING FROM (ablative-based possession)
(40) a. bets’į ‘from 3SG’ ‘it’s his/hers, s/he has it’
b. suga bets’į ‘3SG has sugar’ ‘s/he has diabetes’
The disease metaphor in more general use in Dene Sųłiné is an anthropomorphic one. This
metaphor, illustrated in (41), revolves around A CONDITION/DISEASE IS A WILD ANIMAL
metaphor. We know this image of a wild animal is invoked on account of the fact that the verb
stem -dak ‘eat’ is only used with non-domesticated animals.11
DISEASE IS A DEVOURING ANIMAL
(41) a. gu sedak ‘worms are eating me’ ‘I have cancer’**12
b. dekoth sedak ‘phlegm is eating me’ ‘I have a cold’
c. shíratth’en sedak ‘heartburn is eating me’ ‘I have heartburn**’
d. ya sedak ‘lice are eating me’ ‘I have/am infested with lice’
11 For example, -dak is not used with dogs (the same verb stem associated with human eating is used
instead) and it can be used figuratively with humans.
i. nunie deni hedak ‘the wolf is mouthing a moose’ ‘the wolf is eating a moose’
ii. łį {*hedak / shetį} ‘the dog is *mouthing/eating its food’ ‘the dog is eating’
iii. betsąkwié hedak ‘he’s mouthing his wife’ ‘he’s beating his wife’
12 Hotze Rullmann (personal communication) reminded me that in English, the disease, cancer, is named
for a crab that pinches and bites––an underlying image similar to the Dene Sųłiné case.

21
Not surprisingly, many other disease and physiological state expressions are built around body
part terms. The body part, as the actual or supposed locus of the physiological response or
condition, figures prominently in the following expressions, many of which are metaphorical and
metonymic simultaneously. They tend to feature both a PART FOR WHOLE metonymy and a
SYMPTOM IS DISEASE/CONDITION metaphor, as shown in (42). The ailments, expressed via verbs
of hurting, being swollen, or stopping, can be temporary or chronic.
PART FOR WHOLE and SYMPTOM FOR DISEASE/CONDITION
(42) a. sedzídithé hílgház ‘my lungs, they’re swollen’ ‘I have pneumonia’
b. sedzídithé eyá ‘my lungs hurt’ ‘I have tuberculosis’
c. sedząghe eyá ‘my chest hurts’ ‘I have bronchitis’
d. sebie eyá ‘my stomach insides hurt’ ‘I’ve got diarrhea’
e. setthí eyá ‘my head hurts’ ‘I’ve got a headache’
f. sedheri eyá ‘my liver hurts’ ‘I have cirrhosis’
g. bedzie hįʔą ‘3SG heart stopped’ ‘s/he had a heart attack’**
I list a few miscellaneous expressions in (43) which defy neat classification, although they are
clearly figurative. Whereas eyá ‘it hurts’ is a common expression covering a variety of ailments
(and even psychological states; cf. (50g)), da ‘disease’ has also given rise to a number of
GENERIC FOR SPECIFIC metonymies:
GENERIC FOR SPECIFIC
(43) a. dadhi ‘disease moves/happens’ ‘suffering’
b. dene dadhihi ‘person who suffers’ ‘invalid’
c. łį dádaé ‘dog mouth-disease’ ‘rabies’
Beyond disease proper, a number of less severe ailments––skin conditions, in particular––are
expressed via a figurative lexicalization, as shown in (44). In the examples in (44c-f), tth’i is
another form for ‘body’, unrelated to -tthén encountered in (25).
SYMPTOM IS DISEASE or CAUSE FOR EFFECT and WHOLE FOR PART
(44) a. bek’erek’os ‘it’s red on 3SG ‘s/he’s got measles’
b. dene tthí nédheli ‘person’s head, that which is hot’ ‘fever’ [cf. (74b)]
c. dene tth’i slini ‘(on) person’s body it is evil’ ‘hives’
d. dene tth’i háretł’ézi ‘(on) person’s body that which is blue’ ‘bruise’
e. dene tth’i dík’ǫ((si ‘(on) person’s body that which is red’ ‘rash’
f. del níłtth’éli ‘blood that flows together’ ‘blood clot’ [cf. (39f)]
3.2.4 Physiological states and conditions, both permanent and ephemeral
Many physiological conditions have lexicalized from more literal expressions, some of which
describe the underlying cause of the condition, particular symptoms, or the result of the

22
condition. Some of these are built metonymically around a body part term. Others translate a
specific behavior into a general condition. In (45) are expressions for being able to see or not (i.e.
being blind) and for being able to hear or not (i.e. being deaf). Notably, the same verb stem, -t’į,
is used for ‘seeing’ and ‘hearing’. This stem is usually glossed as intransitive ‘see’, but perhaps it
would be better to gloss it more generally as ‘perceive’. The predications for seeing and not
seeing in (45a-b) and (46a) as well as the counterpart predications for hearing and not hearing in
(45c-d) and (46b) all contain morphemes that appear to be incorporated body parts––na(ghe)-
‘eye’ and dzi(ye)- ‘ear’––not an uncommon verb formation strategy in Athapaskan (cf. S. Rice
2009: 124-126 and passim for discussion about Dene eating expressions incorporating ‘mouth’
terms). Indeed, the examples in (46c-d), including those in (47), also feature a body part as a
salient participant, if not the logical subject.
PART FOR WHOLE and (DYS)FUNCTION FOR CONDITION
(45) a. nast’į ‘I eye-see’ ‘I see well’
b. nast’įle ‘I eye-see-not’ ‘I can’t see/I’m blind’
c. dzióst’į (< dziyé-hó-st’į) ‘I see in my ear’ ‘I can hear’
d. dziyóst’įle (< dziyéhóst’į) ‘I see in my ear-not’ ‘I can’t hear/I’m deaf’
PART FOR WHOLE and ABSENCE IS DYSFUNCTION
(46) a. naghedį ‘eyes-without’ ‘blindness’
b. dziedį [< dziye-dį] ‘ears-without’ ‘deafness’
c. dziyédįhi ‘ears-without-NMLZ’ ‘deaf person’
d. dzádį ‘legs-without’ ‘crippled/paralysis’
e. beyatié húle dene ‘3SG talk is gone person’ ‘deaf-mute’
PART FOR WHOLE and RESULT FOR CAUSE
(47) a. sedzie hįdhų ‘my heart is numb’ ‘I’m hungry’
b. sedeyaghe hega ‘in my throat, it is dry’ ‘I’m thirsty’
c. bedhenįt’i ‘3SG’S throat it is tight’ ‘s/he’s a glutton’
d. seyidaniʔa ‘my breath is blocked’ ‘I have a lump in my throat’**
Dene Sųłiné features a few other physiological state predications that are constructed on the basis
of a general SYMPTOM FOR CONDITION or RESULT FOR CAUSE metonymy, shown in (48). The
examples in (48b-c) also feature an orientation metaphor where OUT/AWAY IS NEGATIVE.
SYMPTOM FOR CONDITION (48a) and RESULT FOR CAUSE (48b-c)
(48) a. heldok ‘3SG is cramping/convulsing’ ‘s/he’s an epileptic’
b. ch’a ahostį ‘my appearance is away/off-putting’ ‘I’m ugly’
c. dene ch’aríldhen ‘person thinks/moves away’ ‘s/he’s deformed’
In addition, there are physiological and emotional state predications in that are constructed on the
basis of a particular metaphor within a general CAUSE FOR RESULT metonymy. Specifically, the

23
targeted state is conceptualized as an entity by means of an ONTOLOGICAL metaphor. Moreover,
the now reified condition is both the subject of the expression and metaphorically construed as
being physically located on or overcoming the experiencer, who by virtue of an OBJECT FOR
EXPERIENCER metonymy is coded as the location. The examples given in (49) are highly
conventionalized. In fact, (49a) with an unspecified subject, generally means ‘to be drunk.’
BEING ON IS HAVING EFFECT ON and UNDERGOING/SUFFERING IS BEING
(49) a. sek’įdher ‘UNSPEC is happening on me’ ‘I’m drunk’
b. dáda sek’įdher ‘illness is happening on me’ ‘I’m sick’
c. beł sek’įdher ‘sleep is happening on me’ ‘I’m sleepy’
d. dekoth sek’įdher ‘phlegm is happening on me’ ‘I’m getting a cold’
e. ts’udi sek’enádher ‘laziness is on top on me’ ‘I’m feeling lazy’
f. sek’enádher ‘UNSPEC is on me’ ‘it’s bothering me’
3.2.5 Psychological State Predications
As was the case with many disease and physiological state expressions in which some body part
stood as the actual or supposed locus of the physiological cause or symptom, so too are many
attitudinal or psychological state predications based on a BODY PART FOR WHOLE PERSON
metonymy whereby the predicated (literal) condition of the body part suggests a figurative
condition of the whole person. As in many languages, the heart is construed as the locus of
emotion in Dene Sųłiné. However, the figurative interpretation of specific attributes of hearts is
somewhat different between it and English. The examples in (50) illustrate how an attribute of a
part (heart) stands for the attitude of the whole (person), while those in (51) show how the
behavior of the heart stands for some emotional reaction by the person.
ATTRIBUTE OF PART (heart) FOR ATTITUDE OF WHOLE (person)
(50) bedzie… ‘her/his heart…’ ‘s/he’s…’
a. …netł’edh ‘…is mighty/powerful’ ‘…hopeful, determined’
b. …nátser ‘…is strong’ ‘…courageous’
c. …denur ‘…is soft’ ‘…kind/humble’
d. …nezǫ ‘…is good’ ‘…good’
e. …necho ‘…is big’ ‘…big-hearted’**
f. …hule ‘…is absent’ ‘…heartless’**
g. …eya ‘…hurts’ ‘…heart-broken’** [cf. (42g)]
BEHAVIOR OF PART (heart) FOR EMOTIONAL REACTION OF WHOLE (person)
(51) a. bedzie nałther ‘3SG’s heart shivered/shook’ ‘s/he was startled’
b. bedzerełtth’er ‘3SG’s heart took off’ ‘s/he got scared’
‘her/his heart was racing’**
c. yets’én bedzie nághídá ‘3SG’s heart moves towards 4SG’ ‘s/he’s furious at him/her’
d. sedzie t’a henesłį ‘I’m happy because of my heart’ ‘I’m elated’ (very positive)

24
Most of the remaining figurative expressions in my corpus pertaining to emotions, attitudes, or
behaviors involve metaphors within metonymies. Generally, they are of two types. In (52), I give
examples of chained metonymies whereby a CONTAINER (-ni- ‘mind’) STANDS FOR ITS
CONTENTS (thoughts) and the CONTENTS STAND FOR THE EFFECT THEY PRODUCE. In addition,
in (52) and (53a-c), orientation metaphors are invoked, whereby IN IS POSITIVE emotionally,
while OUT/AWAY/ABSENT IS NEGATIVE. The examples in (53) and (54) all involve the same
generic verb stem of ideation {-dhen/-then}. Through juxtaposition or incorporation of a PP or
NP complement, specific types of mental activities or states can be predicated. In (53d-e) and
(54), the specific target of ideation––the PP or NP complement––stands for a general attitude or
habitual behavior in a SPECIFIC IS GENERIC metonymy, at the same time that the target of
ideation becomes the target of emotion in a THOUGHTS ARE EMOTIONS metaphor.
(52) a. sinié (< sini yé) ‘(it’s) in my mind’ ‘I’m happy/glad/pleased’
b. sįnik’éch’a ‘(it’s) away from my mind’ ‘I’m disappointed’
c. ą(nihiʔá ‘my mind is in the wild’ ‘I’m lonesome’13
d. ʔenilé (< ʔeni ʔilé) ‘3SG is not minded’ ‘s/he’s naughty/silly’
(53) a. ch'a nįdhen ‘3SG thinks away’ ‘s/he’s stubborn’
b. bech’a nesthen ‘I think away from 3SG’ ‘I disagree with him/her’
c. behéł nesthen ‘I think with 3SG’ ‘I agree with him/her’
d. beka nesthen ‘I think for 3SG’ ‘I want him/her/it’
e. tsąba ghą nįdhen ‘3SG thinks about money’ ‘s/he is miserly’
(54) a. sąnesthen ‘I play-think’ ‘I’m good with children’
b. dlonesthen ‘I laughter-think’ ‘I have a good sense of humor’
c. estenesthen ‘I grief-think’ ‘I’m feeling sorry for myself’
Other generic verbs of cognition work similarly in Dene Sųłiné. By means of an incorporated
postpositional or nominal complement, a specific mental state, attitude, or emotional reaction has
clearly become lexicalized. Many of these formatives are no longer productive.14 Examples of
GENERIC FOR SPECIFIC craving and thinking are given in (55)-(56):
(55) a. bérbaidher ‘for meat I suffer/crave’ ‘I’m hungry’
b. túbaidher ‘for water I suffer/crave’ ‘I’m thirsty’
c. ʔebaidher ‘for unspecified I suffer/crave’ ‘I’m horny’
d. *tsąbabaidher ‘for money I suffer/crave’ *‘I’m greedy’
13 Compare (52c) with the incorporated -ni- ‘mind’ as the logical subject, with the less partonymic ą(hiya,
lit. ‘I went into the wilderness’, fig. ‘I’m lost’ (literally or figuratively). The same orientation metaphor
obtains, since being ‘out there’ is equally construed in a negative way.
14 Their productivity is a topic addressed in Rice (1997).

25
(56) a. bet’a néthiʔa ‘I cognize with 3SG’ ‘I’m {devoted to, trust} him/her’
b. bena néthiʔa ‘I cognize against 3SG’ ‘I’m vengeful against him/her’
Finally, there are a handful of expressions in my corpus that describe an attribute or attitude in
terms of a particular activity. Specifically, they build on the idea that doing something well or
not well implies a general attitude or attribute surrounding the process. These are therefore PART
FOR WHOLE metonymies in the sense that something done well (or not) is something done
habitually. Moreover, (57c-d) also involve a PERCEPTION IS COGNITION metaphor.
(57) a. dene hotie ghęna ‘3SG is living well’ ‘s/he’s healthy’
b. dene hǫzǫ yałti ‘3SG prays good’ ‘s/he’s religious’
c. suwareltį [<sugha ghereltthį] ‘3SG hears well’ ‘s/he understands’
d. sugha ghereltthįle ‘3SG didn’t hear well’ ‘s/he misunderstood’
(58) a. axe eˀerit’į ‘the one who seems proud’ ‘snob’
b. axe ghelkádh ‘3SG is trotting proud/pretty’ ‘s/he’s well off’
3.3 Fauna and their feathers, fur, fins, and feelers
An extremely robust lexical field for metonymies involves animal naming. Folk etymologies
abound for explanations behind some of the lexicalizations presented here. We need not be
particularly interested in the details here, except to say that some aspect, attribute, physical
property, behavior, salient association, or the like has been codified for referential purposes.
Consequently, metonymy seems to be the dominant semantic trope, relativization the usual
morphological mechanism, and mythical beliefs, as much as physical attributes, habitual
behaviors, or location associations, the source of the particular metonymy.
3.3.1 Naming mammals
As I stated earlier, figurativity in Athapaskan nomenclature is not limited to terms of
acculturation. Nowhere is this observation more readily apparent than in terms for animals in
Dene Sųłiné. Both indigenous and introduced fauna are named via metonymies and metaphors,
as only a dozen or so native fauna are named with monosyllabic stems.15 Superordinate terms are
15 Here is a nearly exhaustive list of monolexical animal terms: chedh ‘duck’, dą ‘mole’, deł ‘crane’, dih
‘grouse’, dza ‘dove’, dzen ‘muskrat’, gah ‘rabbit’, hah ‘goose’, łį ‘dog’, łué ‘fish’, sas ‘bear’, tha
‘marten’, tsá ‘beaver’, and ts’i ‘porcupine’. It fails to include the three animals of most critical food value
to northern Dene peoples, the caribou, the moose, and the bison (buffalo). I address the ‘bison’ term in the
text. The word for caribou, etthén, is not mono-lexical. It could derive from e-tthén (lit. ‘its living flesh).
Some bands of Dene Sųłiné were in fact referred to as ‘caribou eaters’ and it would not be surprising if
either the flesh word came to be associated specifically with caribou (like O.E dēor ‘ruminant animal’ →
‘deer’) or the caribou term, as the most common food source, generalized to include all edible flesh
sources. As for the ‘moose’ term, dení (or deníe) it seems constructed out of a common thematic prefix

26
rare in Athapaskan languages, but a few exist, although they are metonymic, based on either a
salient PART or a salient BEHAVIOR standing FOR THE WHOLE, as shown in (59):
(59) a. ech’erisline ‘the evil-tendoned one’ ‘wild animal’
b. tich’anadíe ‘the one that wanders outside’ ‘wild animal’
One could argue that the examples in (59) also feature a GENERIC FOR SPECIFIC metonymy since
all animals have tails and all animals wander. This trope is more obvious in the next set featuring
more specific indigenous animals in the Dene world:
GENERIC FOR SPECIFIC & ATTRIBUTE FOR ENTITY
(60) a. nągídhi ‘the one who twists’ ‘fox’
b. ną(mbie ‘the one who swims across’ ‘otter’
c. nadudhi ‘the one who slithers’ ‘snake’
d. ną(ghai ‘the one who takes down’ ‘wolverine’
e. nunie ‘nostril to the ground’ ‘wolf’
f. nųniets’elas ‘little wolf (nostril-to-the-ground)’ ‘coyote’
g. nultsįhi ‘the one who sniffs the ground’ ‘skunk’
h. tełk’ali ‘clean sock one’ ‘weasel’
i. ejere ‘the stinky/rotten one’ ‘bovine’ (bison, cow)
A small set of indigenous animals is lexicalized using some sort of marking reversal or
metaphorical extension of a native animal term. Although (60i) originally meant ‘bison’, it has
now come to refer only to domesticated cattle. Bison are now referred to as tłoghįjere ‘grass-
cow’. Similarly, in some Dene languages, the widely cognate term for ‘dog’, {tłį, łį, tłe}, is now
used exclusively for ‘horse’ rather than the common label––something like łįcho ‘dog-big’. This
has required a different lexicalization strategy for the old native concept, ‘dog’, which inevitably
comes about through a behavioral attribute metonymy like those in (60). The only indigenous
animal term I know of that is clearly metaphoric in Dene Sųłiné is the word for ‘bat’. It involves
the modification of a native animal term, tsá ‘beaver’, resulting in an intra-domain metaphor. In
addition, as is common with many introduced (via colonization) or encountered (via migration)
animal terms, it is lexicalized like the others in (61), through the addition of a quantity- (size) or
quality- (valuation) based suffix, or both. I list a host of similar examples from multiple semantic
fields in Tables 1-3 in §3.5.3.
(61) a. tsáret’anaze ‘little flying beaver’ ‘bat’
b. łįcho ‘big dog’ ‘horse’
c. etthénslinaze ‘little evil caribou’ ‘sheep, lamb, goat’
de- found with stative verbs, as discussed in (6), and the generic stem for ‘food’ ni. So dení, might
literally be glossed as ‘that which is food.’ Moose are treated linguistically like human beings, unlike
other animals, as they both eat and walk with the same verb stems used for human action. See fn. 5.

27
Generally, all introduced or encountered fauna have fanciful (i.e. metaphorical) names, as
predicted by Brown (1999). The expressions in (62) exemplify typical lexicalization strategies
deployed, one involving a GENERIC ATTRIBUTE FOR SPECIFIC ENTITY metonymy and the other
an intra-domain, BIOTYPE-FOR-OTHER-BIOTYPE metaphor in an attribute-based metonymy.
(62) a. dene hedarelį ‘the one who imitates a man’ ‘monkey’
b. gu detth’eni ‘the worm/bug that’s boned’ ‘turtle’ [cf. (63d)]
Because terms for most introduced animal concepts for non-indigenous animals like ‘elephant’,
‘giraffe’, refer to the obviously salient body part (i.e. trunk or neck) and are both rarely used and
have not conventionalized across speakers or communities, I refrain from including them here.
3.3.2 Naming non-mammals
Due to the paucity of figurative examples, I lump metaphoric and metonymic terms for insects
(63), birds (64), and fish (65) together in this section. Most of these examples are metonymic,
involving the expression of some attribute or behavior which modifies some generic insect term
like gu ‘worm/bug’, yá ‘flea/louse’, or tł’izi ‘horsefly’, as shown in (63a-g). Most of these are
compounds, which generally involve some sort of metaphorical marking reversal or other
semantic extension. The remaining examples in (63h-k) are deverbal and therefore metonymic.
INSECT TERMS
(63) a. k’áígúé ‘willow worm’ ‘caterpillar’
b. gukale ‘flat worm; worm that is flat’ ‘bedbug’**
c. guslinaze ‘little evil worm’ ‘grub’
d. gu detth’enaze ‘the little boned bug’ ‘beetle’ [cf. (62b)]
e. yagolas ‘little sky worm’ ‘butterfly’**
f. tthot’įné yá ‘Englishman’s louse’ ‘flea’
g. tł’izi tthoghe ‘yellow horsefly’ ‘bee’
h. dejoli ‘the one that’s pointed’ ‘mosquito’
i. horádzi ‘the one who winds around’ ‘spider’
j. honeltónas ‘little holdable thing’ ‘no-see-ums’**
k. chądii ‘the ones that travel’ ‘ants’
Dene Sųłiné has two generic ‘bird’ terms, one for raptors, det’ani, and the other for smaller
(usually) songbirds, įyese, which is unanalyzable to my speakers, though clearly composed. Both
terms are metonymic and enter into complex expressions for naming specific types of birds.
(64) a. det’ani ‘the winged/feathered one’ ‘bird (especially a raptor)’
b. dádzení ‘the black-beaked one’ ‘loon’
c. tł’oghetsáné ‘grass-jay’ ‘blackbird’

28
d. įyese (unanalyzable) ‘(song) bird’
e. dení įyesaze ‘little moose bird’ ‘hummingbird’
f. įyese yałtihi ‘the bird that talks’ ‘parrot’
Łué is the generic Dene Sųłiné word for ‘fish’, although it is probably derived from the most
prototypical northern fish, łú ‘whitefish’. In (65), we see some simple examples based on the
stem łué ‘fish’. I survey some specific examples of what are in fact metonymies based on
qualitative properties of entities.
(65) a. łú ‘whitefish’
b. łué ‘of/like whitefish’ ‘fish’
c. łuezáné ‘fish-black’ ‘trout’
d. łuezáné łą(t’i ‘it’s like trout’ ‘salmon’
e. sąt’íe (< zan/zen-t’í) ‘black-it.is/appears’ ‘grayling (fish)’
f. egóthécháe (> gócháe) ‘its neck is big’ ‘sucker (fish)’
g. déldeli ‘the red-colored one’ ‘red sucker’
h. echui ‘the spiny one’ ‘pickerel’
3.4 Places and spaces
Topographic and geographic terms are relatively under-represented in my corpus. Place names,
on the other hand, are nearly always metonymic, with an important event or geological feature
entering into the name. I do not have space to list any here, most of which only name small
hamlets, rivers, and lakes particular to the Dene Sųłiné world. Rather, I list figurative expressions
that describe natural phenomena (§3.4.1) and cultivated places (§3.4.2).
3.4.1 The natural world
Of the classic four basic elements, ‘earth’ (nih), ‘sky’ (ya), ‘fire’ (kón), and ‘water’ (tu), all are
monomorphemic and unanalyzable. However, all four––along with monosyllabic terms for other
natural entities such as ‘sun’ (sa), ‘star’ (tthén), rock’ (tthe), ‘sand’ (tthai), ‘rain’ (chą), ‘falling
snow’ (tsįł), ‘snow on the ground’ (yath), ‘ice’ (ten), ‘river’ (des), ‘hill’ (sheth), ‘cover’ (tél)
and ‘island’ (nu)––enter into composite expressions which are figurative in some way.
METAPHORICAL COMPOUNDS
(66) a. shéth-chogh ‘hill-AUG’ ‘mountain’
b. nįn-teli ‘earth-cover’ ‘muskeg’
c. tł’o-teli ‘grass-cover’ ‘prairie’
d. sa tł’ulé ‘sun rope’ ‘sunbeam/moonbeam’
e. tetł’e-zaé ‘night-sun’ ‘moon’
f. tsįł-lu ‘frozen falling snow’ ‘hail’

29
METONYMIC DEVERBAL EXPRESSIONS
(67) a. yéłką (<ya héłką) ‘sky-it lights up’ ‘it’s dawn’
b. hoye ‘inside AREAL (a place)’ ‘hole’
c. xátaįlį ‘water which is flowing out’ ‘spring’
d. nįłts’i ‘that which blows ‘wind’
e. horádzi yélu ‘the spider weaved it’ ‘rainbow’**
f. náhagez ‘PL things move randomly/stir’ ‘fog’
g. nayelka nághegez ‘it lights it up, it stirs’ ‘northern lights’
3.4.2 The human world
As prototypical hunter-gatherers, Dene peoples were traditionally mobile, engaging in seasonal
migrations as a band as well as more solitary living out in the bush along a trap line. Most
encampments were temporary and permanent settlements only came about with the signing of
treaties with the British Crown in the late 19th century. The notion of a human place centers
around the concept ‘fire’, kón. The ‘place of the fire’ (lit. ‘on the fire’) is kónk’é. Being
‘around/at/beside the fire’ is kóní/kóné. I believe that there is good evidence for proposing that,
over time, this PP reduced phonetically as it became more abstract semantically. The term kǫ(ę
(many alternate spellings exist, including kóę, kǫ(é, and even kǫ(ę) has taken on a variety of
contemporary meanings from ‘campsite’, ‘house’, or ‘building’ all the way to ‘village’ or ‘town’.
Thus, ‘(the place) beside the fire’ extended to mean ‘(the place) where people are’ in what is
perhaps a WHOLE FOR PART metonymy. This highly productive term has compounded
morphologically with a large number of lexical items to produce many of the specialized
buildings (or rooms) that one finds in a township. Semantically, these composite terms involve a
CONTENTS FOR CONTAINER metonymy, as happens when a space turns into a place.
CONTENTS FOR CONTAINER
(68) a. yatikǫ(ę ‘talk/prayer-house’ ‘church’
b. tsąbakǫ(ę ‘money-house’ ‘bank’
c. tsąkǫ(ę ‘shit-house’ ‘toilet, outhouse’
d. eyakǫ(ę ‘hurt-house’ ‘hospital’
e. ejerekǫ(ę ‘cow-house’ ‘barn’
f. dzółkǫę ‘ball-house’ ‘pool hall’
g. sekwi hóneltenikǫ(ę ‘children they.are.taught-house’ ‘school’
Traditional dwellings of Northern Dene peoples were tipis––nįbáli (literally, ‘that which
hangs/flaps/drapes), erected using multiple animal hides sewn together and wrapped around a
dozen or so long wooden poles arranged in a conical shape. The term has come to mean any
canvas tent, a (weak) metaphorical extension of this original metonymy. The same term can
mean ‘sails’ as well, as in ts’i-nįbáli ts’i ‘sailboat’ (lit. ‘boat-canvas boat’).

30
Two very productive place-building morphemes which are suspiciously cognate (and for
which a plausible shared etymology could be projected), which overlap semantically, and for
which any tone differences are usually neutralized, are -k’e ‘on/place’ and -k’é ‘on surface
of/hole’. Despite inconsistencies in how speaker-consultants pronounce these morphemes (which
I have tried to transcribe and represent orthographically as uttered), speakers are fairly adamant
about glossing one as a ‘place’ and the other as a kind of ‘hole’. Significantly, there is a nice
minimal pair between (69b) and (70b) which confirms the meaning contrast. In any case, the
semantic extension from an adpositional relation to an abstract location is well attested cross-
linguistically. The Dene Sųłiné postposition, -k’e, implies contact between some unspecified
figure and a surface-like ground location. A metonymic shift to the actual place that the contact
surface occupies does not require much cognitive imagination, nor does a further extension from
contact with two-dimensional surface to penetration through it in order to access a three-
dimensional container or hole below or behind it.
(69) a. k’e ‘on’ ‘place’
b. yak’e ‘sky-place’ ‘heaven’
c. kónk’é ‘fire-place’ ‘campsite’
d. níhók’é ‘land-place’ ‘farm, field’
e. náíník’é ‘back-and-forth place’ ‘store’
f. tth’áík’é ‘dish-place’ ‘cupboard’
(70) a. k’é ‘on’ ‘hole, opening’
b. yak’é ‘sky-hole’ ‘window’
c. tuk’é ‘water-hole’ ‘well’
d. yoréldedhék’é ‘key-hole’ ‘lock’
e. bąlaghek’é ‘button-hole’ ‘buttonhole’
3.5 Terms of acculturation
Culture drives lexicalization whenever there is a need for increased referential diversity as
happens through experiential circumstance or cross-cultural contact. Lexical innovations brought
about by the latter are the most striking because there is generally a high expectation that foreign
concepts will be lexicalized in a creative and figurative way. While there are several dozen
borrowed terms (from French, chiefly) for physical artifacts, domesticated animals, church-
related concepts, and the like introduced into Dene culture by Europeans, most terms of
acculturation have been constructed using morphological material and lexicalization patterns
already present in the language and highly pervasive in Athapaskan languages generally. With
the exception of some wholesale conversions already discussed (e.g., -la, ‘hand’ and -ke ‘foot’),

31
there are few straightforward extensions that do not involve additional modification through
compounding or relativization. Some notable examples are given in (71):
NATURAL IS ARTIFACTUAL (71a-b) and FORM/FUNCTIONAL SIMILARITY (71a-c)
(71) a. sa ‘sun’ ‘clock’
b. tthe ‘stone’ ‘pipe, cast’
c. įtsółé ‘rosehip’ ‘tomato’
Each of the examples in (71) is based on a metaphor. Although I have not performed a
quantificational analysis like Brown (1999) did for his 77 terms of acculturation in Dene Sųłiné
to see whether, in fact, natural kinds tend to be lexicalized metaphorically while artifacts tend to
be metonymic, this is the impression I have from my study of the language. Most of the terms
described and illustrated in this section involve artifacts and concepts that would have been
commonly found in Dene homes and communities through the 1940s-1950s, the last time a
majority of the community still spoke Dene Sųłiné as a first language. The ravages of the
residential school system and land loss to the government, military, oil and gas industry, and
non-native farmers really took their toll on family structure and traditional livelihoods after
World War II. Therefore, most terms of acculturation listed here have a 19th century to mid-20th
century feel to them.
Many of the natural and acculturated artifact examples in (72)-(95) involve metaphors
(based on physical and/or functional similarities) and metonymies simultaneously. I have loosely
grouped the examples by lexical field or morphological stem, depending on where the highest-
level generalization can be made. For the most part, the examples in this extended section
involve the lexicalization of nominal entities. These may come about through compounding,
deverbalization, or suffixation. The end result is a new set of nouns (and a few stative relations)
from an old set of diverse lexical material.
3.5.1 Recycling the old for the new
Compounds with -tu ‘water’ are especially numerous, still productive, and a good place to start
in describing Dene Sųłiné terms of acculturation. The contribution of -tué as the head element in
the compounds in (72) is likely to contribute the notion of ‘liquid’, in a SPECIFIC FOR GENERIC
initial metonymy. The modifying element suggests the source domain to which the thing named
by the overall compound belongs (perhaps with the exception of kóntué ‘alcohol’ (lit. ‘fire-
water’) and k’estué ‘watermelon’ (lit. ‘aspen-water’ → ‘sap’). Although these two exceptions are
strikingly metaphorical, the remaining examples in (72) could safely be described as metonymic
in, ironically, a GENERIC FOR SPECIFIC way.

32
(72) a. tué ‘(of) water’ ‘liquid’
b. jiétué ‘berry-water’ ‘wine, juice’
c. tł’olátué ‘grain-water’ ‘beer’
d. senaghetué ‘my eye-water’ ‘my tears’
e. dechentué ‘tree-water’ ‘sap’
f. k’estué ‘aspen-water’ ‘watermelon’
g. eghézétué ‘egg-water’ ‘egg white’
h. erihtł’ístué ‘writing-water’ ‘ink’
The contribution of tu- ‘water’ as the modifying element in the compounds in (73) contributes,
respectively and metonymically, the notion of ‘constituency’ in (a), ‘typical contents’ in (b), and
‘location’ in (c)––all subparts or associated features of water, rather than the whole substance.
(73) a. tutłesé ‘water-grease’ ‘kerosene’
b. tuteli ‘water-container’ ‘bottle, pail’
c. tułué ‘water-fish’ ‘salmon’
In (74), the contribution of tu- as an incorporated verb complement or object of an incorporated
postposition only invokes literal water (arguably) in (74c-e). As all the examples involve
processes standing either for entities (74a-b) or other processes (74c-e), they are all metonymic.
All five examples suggest the initial phase of the process that brings about the lexicalized result.
(74) a. tuniłkedh ‘water comes together’ ‘blister’
b. tu nédheli ‘water which is hot/heated’ ‘soup’ [cf. (44b)]
c. túbaidher ‘for water I crave’ ‘I’m thirsty’
d. túyenasther ‘I’m in water’ ‘I’m swimming’
e. tú nédąn ‘3SG drank water’ ‘s/he drowned’
As in pre-colonial times, wood (dechen), stone (tthe), and rawhide (-dheth) were the most
common source materials used by Indigenous peoples in the subarctic to construct cultural
artifacts for a post-colonial life-style. The following domestic objects could be considered double
metonymies. They are compounds, of a sort, with the stem (left-most element) describing the
source material, wood, or the shape of the new artifact, stick. In (75a-d), the modifying element
describes the process that the artifact is associated with (sleeping, sitting, writing), while in (75e-
g), it describes a salient entity associated with some process. Moreover, in these last four
examples it is shape (stick-like) rather than internal constituency that is most likely to be the
profiled contribution of the stem -chené. Nevertheless, I have glossed all uses as ‘wood’.
(75) a. etéchené ‘one person lies-wood’ ‘bed’
b. edáchené ‘one person sits-wood’ ‘chair’
c. łįchogh k’e dáhchené ‘one person sits on horse-wood’ ‘saddle’

33
d. erihtł’íschené ‘one marks it-wood’ ‘pencil, pen’
e. edhéthchené ‘hide-wood’ ‘stretcher (frame for hide)’
f. hodethchené ‘screw-wood’ ‘screw-driver’
g. jéthchené ‘fishhook-wood’ ‘fishing pole’
The examples in (76)-(77) involve figurative extensions around the morpheme yú ‘clothing,
clothes’. In (76), we find metonymic compounds built on the same essential and general
metaphor whereby the most personal of personal effects––clothing––generalizes to mean ‘stuff,
belongings’ so it can then particularize through a metonymy to mean different kinds of personal
effects. In (77), we find other household items that pertain to storing or cleaning clothes. Some
of these are what Brown (1999) would call descriptive comments about functional utility.
GENERIC FOR SPECIFIC
(76) a. yúé ‘of clothing, clothes’ ‘belongings, gear’
b. layúé ‘hand-gear’ ‘tool, equipment’
c. jéth layúé ‘fishhook-gear’ ‘fishing tackle’
d. sąyúé ‘fun-gear’ ‘toy’
e. yísį hoyúé ‘inside it place-gear’ ‘furniture’
f. sǫyúé ‘good-gear’ ‘jewelry’
(77) a. yúteli ‘clothes-container’ ‘wash tub’
b. yú dechen teli ‘clothes-wooden-container’ ‘trunk’
c. yúch’elaze ‘little torn cloth/clothes’ ‘rag’
d. bet’á yú delk’ali ‘with it, that which makes clothes white’ ‘bleach’
e. bet’á yú k’enáłtsiłi ‘with it, that which makes clothes clean’ ‘washing machine’
f. beyé yú thelai ‘in it, one puts clothes’ ‘dresser, suitcase’
In (77e) above, we see that the expression for ‘washing machine’ be-t’á yú k’e-náł-tsił-i (lit.
‘with-it clothes on-it.makes-clean-the one’ or ‘that which makes clothes clean’), features the
verb stem -tsił. This morpheme is cognate with the noun for ‘falling snow’, first seen in (66f) as
well as in the term for ‘snowflake’ tsiłkáré (lit. ‘falling.snow-flat’). No doubt through a form-
based (and perhaps even color-based) metaphor, the morpheme for snow came to mean ‘soap’
(or more accurately, ‘soapflakes’) and, through an INSTRUMENT FOR PROCESS metonymy, the
verb stem for ‘clean’. We find this ‘snow’ → ‘soap’ → ‘clean’ morpheme tsił in a number of
expressions, which, considered individually, might otherwise not seem very figurative:
(78) a. datsił ‘soap’
b. selák’e nástsił ‘I am washing (on) my hands’
c. bet’a eghú k’enátsiłi ‘with it, the thing that teeth are cleaned’ ‘toothpaste’
d. beyé k’enáts’eltsiłi ‘in it, the thing that people are cleaned’ ‘bathtub’

34
Another metaphorically driven polysemy chain arguably holds between the words for
‘sock/stocking’ (the traditional rawhide legging which covered the foot and leg), tel, and the all-
purpose ‘container’ word, teli. Although the latter is considered monomorphemic by speakers, it
has the form of a derived word like tel-li ‘sock-like’ or tel-e ‘of a sock’. Bi-syllabic stems are
exceedingly rare in Dene Sųłiné. In any case, teli, is a very productive stem and enters into
lexicalizations for all manner of closed or semi-closed containers. The examples in (79a-f)
involve a CONSTITUENT-BASED metonymy; the material substance describes the thing it makes.
The examples in (79g-h) involve a CONTENTS FOR CONTAINER metonymy. All the examples in
(79) are, therefore, PART FOR WHOLE metonymies of one sort or another.
(79) a. teli ‘container’ ‘pail, pan, pot, motor’
b. xąįteli ‘root-container’ ‘(spruce root) basket’
c. k’áíteli ‘willow-container’ ‘(willow) basket’
d. tsątsánételi ‘metal-container’ ‘tin can’
e. ttheteli ‘stone-container’ ‘jug’
f. dechenteli ‘wooden-container’ ‘wooden box’
g. tuteli ‘water-container’ ‘bottle’
h. tsąbateli ‘money-container’ ‘cash register, safe’
Like ‘soap’ and ‘container’ examined above, other household items, especially foodstuffs,
have a metaphorical origin in Dene Sųłiné. I list some below in groups with their literal stem(s).
FORM SIMILARITY
(80) a. tł’o(gh) ‘grass’
b. tł’o-lá ‘grass-hand/end’ ‘grain’
c. tł’olá-tué ‘grain-water’ ‘beer’
d. tł’ochenas ‘wooden grass-DIM’ ‘carrot’
(81) a. thai ‘sand’
b. dedhai ‘that which is sandy’ ‘salt’
c. dedhaidzeni ‘salt that is black’ ‘pepper’
FORM SIMILARITY (103a-b) and PROCESS FOR RESULT (103c)
(82) a. łés, leze ‘powder/dust’ ‘flour’
b. konleze ‘fire-powder/dust’ ‘ashes’
b. łést’ɛ(th ‘flour-baked’ ‘bread, bannock’
GENERIC FOR SPECIFIC
(83) a. dene ni ‘human food’ ‘food’
b. dlie ni ‘squirrel food’ ‘nuts, peanut butter’
c. łįcho ni ‘horse (big-dog) food’ ‘oats’
NATURAL FOR ARTIFACTUAL & CONTENTS FOR CONTAINER
(84) a. bekóné ‘its fire’ ‘electricity’
b. betili kóné ‘its container (motor) of fire’ ‘battery’

35
The travois was the traditional device used by pre-contact indigenous peoples of North
America to drag loads of food, clothing, or other belongings over land. It consisted of a few long
wooden poles joined by pieces of sinew that held a pack to be pulled by a human or dog team.
The Dene used this method of transporting goods well into historic times and the travois or
béthchené (lit. ‘pack/load-wood/sticks’) has extended semantically to mean ‘sled, wagon’ and,
with modification, many other vehicles as well. The land-based vehicle terms of acculturation in
(85) are based on the traditional travois term, so in all cases, we have a kind of intra-domain
metaphor at play in addition to the usual GENERIC FOR SPECIFIC metonymies.
(85) a. béthchené ‘pack/load-wooden’ ‘sled, wagon’
b. kón béhchené ‘fire sled’ ‘train’
c. tłesbéchenéchogh ‘gas-sled-AUG’ ‘truck’
Before contact, Dene peoples had a limited set of material resources (rock, wood, birchbark,
grass, hide, sinew, bone) as well as human body parts (foot, thumb) for constructing and
measuring their world. Since number is not marked on nouns in most Dene languages, the
presence of the numerical quantifier reinforces the domain change from the sphere of qualifying
types of objects to the sphere of quantifying sizes of objects or how much they can contain.
(86) a. įłághe béthchené ‘one sled’ ‘one hundred’
b. įłághe dechene ‘one wood/stick’ ‘one yard/one mile’
c. įłághe dechenchogh ‘one big wood/stick’ ‘one thousand’
d. įłághe ttheteli ‘one stone-container’ ‘one gallon’
e. įłághe denechédh ‘one human-thumb’ ‘one inch’
f. įłághe deneké ‘one human-foot’ ‘one foot’**
3.5.2 Modifying the old for the new: Neologisms through qualitative suffixation
A very common lexicalization strategy is to augment a stem (either a bare root or some product
of prior derivation) through a set of “qualitative” suffixes. As mentioned in the introduction to
§3, these encompass a small set of qualifiers, which can either modify their stem or completely
relexify it. Here, we are interested in the latter case. Although the full set includes an
augmentative, several diminutives, a defunctive, gender markers, some shape and color terms, a
marker of high import/valued entity, a marker of low import/rejected entity, and a negativizer, I
will only exemplify a few of these types. I consider them as potentially figurative since they
invoke “scalar” metonymies: the suffix has the effect of partitioning off only one pole of a
dichotomous opposition or only part of a continuous scale (of color, for example) in a PART FOR
WHOLE metonymy. Moreover, in the cases we are interested in, the modified stem refers to a

36
completely different entity (sometimes a different species entirely), suggesting that metaphor is
also responsible for some of the semantic extension. Many of the examples in Table 1 illustrate a
Table 1. Neologisms based on augmentatives or diminutives, or both.
stem
+ AUG: -cho(gh), -chok
+ DIM: -aze
+ DIM: -tsele
łį ‘dog’
łįcho ‘horse’
łįaze ‘puppy’
łįtsele ‘small dog’
dene ‘person, human’
denechogh ‘giant’
dene tselaze ‘dwarf’
dene tselaze ‘dwarf’
chíze ‘lynx’
chízechogh ‘bobcat’, ‘lion’
chizaze ‘cat’
łue ‘fish’
łuechogh ‘whale’
łuaze ‘minnow/smelts’
jíe ‘berry’
jíechogh ‘apple, orange’
jíaze ‘raisin’
háyorįla ‘town’
háyorįlaicho ‘city’
háyorįlaze ‘village’
des ‘river’
desaze ‘creek’
des tsele ‘creek’
datsą( ‘raven’
datsą(ze ‘magpie’
datsą(tsele ‘crow’
nunie ‘wolf’
nunietsele ‘coyote’
yáhtóę ‘deer’
yáhtóę tsele ‘antelope’
łįcho ‘horse’
łįcho tsele ‘pony’
yath-tu ‘fallen.snow-water’
yath-tu-tsele ‘sleet’
yath-lu ‘fallen.snow-frozen’
yath-lu-aze ‘hail’
deneyu ‘man’
deneyuaze ‘boy’
ts’ékwi ‘woman’
ts’ékwaze ‘girl’
sekwi ‘child’
sekwiaze ‘baby’
sekwiaze ‘baby’
sekwazazé ‘doll’
denetthí ‘human head’
denetthíaze ‘postage stamp’
dlíe ‘squirrel’
dlíechoaze ‘gopher’
taretįé ‘rough water’
taretįaze ‘ripple’
kón ‘fire’
kónaze ‘battery’
dzół ‘ball’
dzólaze ‘marble’
nóneshe ‘plant(s)’
nóneshaze ‘vegetables’
tł’oghetsáné lit. ‘grass-jay; fig. ‘blackbird’
tł’oghtsánaze ‘swallow’
béthchené ‘sled/wagon’
béthchenaze ‘bicycle’
tųlu ‘road’
tųluaze ‘path’
jíze ‘jay’ (Western)
jízechogh ‘hawk’
dlúne ‘mouse’
dlúnechogh ‘rat’
det’ani ‘bird’ (raptor)
det’anichogh ‘eagle’
tł’izi tthoghe ‘bee’
tł’izi tthoghéchogh ‘hornet’
erihtł’ís ‘paper, book’
erihtł’íschogh ‘Bible’
teli ‘container’
telichogh ‘barrel’
tł’uk’etį ‘violin’
tł’uk’etįchogh ‘guitar’
tu ‘water’
tuchogh ‘ocean’
ʔih ‘garment’
ʔihchogh ‘parka’, ‘coat’
ke ‘shoe(s)’
kechogh ‘boot(s)’

37
specific type of qualitative metonymy whereby parts of a size scale come to stand for a new (and
different) whole. Lexicalization proceeds via an augmentative suffix or one of two diminutive
suffices or sometimes via a combination of these. Note, too, that the examples in Table 1 run that
gamut from living thing to geographic place to artifact.
Two other less common, but still productive sets of “qualifying” suffixes that have been
used in Dene Sųłiné to build the lexicon are shown in Tables 2 and 3. In the former case, an
intensifier-like suffix, -néthé, with a meaning like ‘big, important’, imparts a high value on the
entity named by the stem, turning it into a different type of entity entirely. In the latter case, the
suffix, -slini/-sline, which is usually glossed as ‘evil’, imparts a low value on the stem concept:
the (new) entity is something to be rejected or feared. The two stems are not really antonyms, but
they occupy positive or negative regions of whatever valuation scale is invoked by the content
domain of the stem. In both cases, significantly, they create new lexical items.
Table 2. Neologisms based on the suffix, -néthé ‘important’.
bare stem
with intensifier -néthé ‘important’
tųlu ‘road’
tųlu néthé ‘highway’
kǫ(e ‘building, town’
kǫ(enethé ‘city’
k’ódheri ‘boss, chief’
k’ódheri néthé ‘king, prime minister’
yałtii ‘priest’
yałtii néthé ‘bishop’
náyałtihi ‘lawyer’
náyałtiinéthé ‘judge’
chą ‘rain’
chą néthé ‘torrent’
des ‘river’
desnethé ‘MacKenzie River’
Table 3. Neologisms based on the suffix, -slini ‘evil’.
bare stem
with dysphemistic -slini/e ‘evil’
tł’o ‘grass’
tł’oslini ‘thistle, weed’
ˀena ‘enemy’ > ‘Cree’
ˀena slini ‘Ojibwe’
ebą(dzaghéjere ‘mushroom’
ebą(dzaghé slini ‘toadstool’
náídíi ‘medicine’
náídíslini ‘poison’
kón ‘fire’
kónsline ‘(Christian) hell’
Other morphological material, from the domains of shape, color, and space, also appear with
nouns as nominal suffixes (87)-(88), post-nominal deverbal modifiers (89), or as postpositional
heads (90), and can lead to lexical innovation. The shape terms, two of which are shown in (87)
and (88), are the simplest and look like the augmentative -chogh and diminutive -aze suffixes
listed in Table 1.

38
(87) a. dekáł ‘it’s flat’
b. t’a kálé ‘the flat one’ ‘pocket’
c. tth’áíkálé ‘dish-flat’ ‘plate’
d. tł’ulekálé ‘rope-flat’ ‘strap’
e. dechenkálé ‘wood-flat’ ‘board, lumber’
f. ts’ikáléchogh ‘boat-flat-AUG’ ‘barge’
g. ts’iłkáré ‘falling.snow-flat’ ‘snowflake’
(88) a. delu ‘it’s cylindrical/sticklike’
b. tenlu ‘ice-cylinder’ ‘icicle’
c. denelu ‘human-cylinder’ ‘corpse’
d. denelué ‘human’s cylinder’ ‘calf of leg’
Color terms16 may be fully deverbal (89a-f) or suffix-like (89g-i) in Dene Sųłiné, probably
depending on the degree of entrenchment or analyzability of the resulting expression. When
modifying a noun, they often create a new lexical item in an ATTRIBUTE FOR ENTITY metonymy.
(89) a. tsątsáné deltthoghi ‘metal that is yellow’ ‘copper’
b. tsąba deltthoghi ‘money that is yellow’ ‘gold’
c. tsąba delgai ‘money that is white’ ‘silver’
d. sas delgai ‘bear that is white’ ‘polar bear’
e. taretį delgai ‘choppy water that is white’ ‘white caps’
f. jíéchogh delzeni ‘big berry that is black’ ‘prune’
g. jíetth’oghé ‘berry-yellow/orange’ ‘orange, corn’
h. jíegaié ‘berry-white’ ‘beans’
i. jíegaié tł’ézé ‘white berry-blue/green ‘peas’
Finally, postpositions can effectively re-lexicalize their nominal complements. We saw many
examples in (69)-(70) with -k’e ‘on’. In (90), I list some additional, yet less robust, examples
with other postpositions or locatives.
(90) a. xáit'ázį( <xáye-t’ázį() ‘winter-behind/against’ ‘autumn’
b. łuk’e ‘fish-on/place/time’ ‘spring’
c. shéth-geze ‘hill-between’ ‘canyon’
16 There are seven basic color terms in Dene Sųłiné. They feature the stative prefix del-, which is strongly
associated with sound and color predications, and an (unanalyzable) stem: delk’os ‘(it’s) red (-colored)’,
delzen ‘(it’s) black’, delgai ‘(it’s) white’, deltses ‘(it’s) brown/faded’, delba ‘(it’s) grey’, and deltthogh
‘(it’s) yellow/orange’. The stems (the last syllable) might be cognate with more interpretable or
transparent stems in other Athapaskan languages, but they remain opaque to modern speakers. An
exception is the term for ‘blue’, which is transparently figurative for most speakers: detłés (lit. ‘(it’s)
grease/lard-colored’). This term can also mean ‘green’. On the other hand, ‘green’ has two periphrastic
exponents, depending on whether it describes––elliptically and metonymically––that which looks like
(the color of) a spruce tree or a leaf. Neither conforms to the usual pattern of color naming; that is, they
lack the prefix del-. On the other hand, they are overt similes (the most rudimentary type of metaphor): el
lą(t’e (lit. ‘it looks like a spruce tree’) or t’ácháí lą(t’e (lit. ‘it looks like a leaf’).

39
d. lá-ʔáné ‘hand-circles around’ ‘ring finger’
e. łés k’e ʔáné ‘on bread-circles around’ ‘bread crust’
f. yeʔáné ‘it-circles around’ ‘wilderness’
g. dene ts’į ʔáné ‘from person-circles around’ ‘offspring’
3.5.3 Seeing the old as the new: Similes
As we gradually move away from compounds and appositive constructions towards more full-
blown periphrastic lexicalizations, we should consider the most literal or explicit of all figurative
tropes––the simile. There are a few conventionalized ones in the language, involving both intra-
domain and inter-domain metaphors.
SIMILES
(91) a. etteláze lą(t’i ‘it looks like a little colon’ ‘sausage’
b. łuezáné lą(t’i ‘it looks like trout’ ‘salmon’
c. shįth lą(t’i ‘it looks like wart’ ‘corn (on feet)’
d. sónibán yúé lą(t’i ‘it looks like silk cloth’ ‘nylon’
3.5.4 Deconstructing processes
In Athapaskan languages, the verb is fully propositional. It generally inflects for subject and
object, valency, tense/aspect/mood, and a variety of thematic and adverbial prefixes. The details
of verbal inflection are complicated and need not concern us here. Pertinent is the fact that verbs
regularly nominalize and relativize (the same process is involved in both cases: suffixation with
the morpheme -(h)i, although it is sometimes incorporated into the stem or is simply absent).
Nevertheless, the result is a fertile source of neologisms that will either be deverbal words that
can enter into further suffixation or compounding or phrasal expressions (in the presence of a
head noun or postpositional phrase). Some of the resulting expressions for artifacts or other
terms of acculturation are syntactically and semantically complex, although they provide fairly
descriptive “functional” labels and thus could be analyzed as process-based metonymies. The
process stands for a participant or location of that process.
PROCESS FOR LOCATION/INSTRUMENT/PATIENT
(92) a. bek’eshíts’elyi ‘on it people eat food’ ‘table’
b. beyághe horétth’ąí ‘that which one hears through it’ ‘radio’
c. benuzeʔį (<benųzí edezelʔį) ‘one looks at oneself through it’ ‘mirror’
d. nak’e ts'ełyai ‘that which lays on the eyes’ ‘glasses’
e. ts'élt'úi ‘smoke/fog which is sucked’ ‘tobacco, cigarette’
(93) a. thenakothi [< theni nakóthi] ‘that which turns alone’ ‘car’, ‘automobile’**
b. ts’ichogh ną(kóthi ‘big boat which turns back & forth’ ‘ferry’
c. ts'ichoretai ‘big boat which flies’ ‘airplane’

40
(94) a. tł’uk’et’į ‘on string, sticklike object is pulled’ ‘violin’
b. tł’uk’et’įchogh ‘big violin’ ‘guitar’
c. tł’iséjeni (<tł’ís héjeni) ‘paper that sings’ ‘accordion’
d. dechentilihéjeni ‘wooden-container that sings’ ‘piano/organ’
(95) a. erihtł’ís ‘it is marked/written’ ‘paper, book, writing’
b. erihtł’ís chené ‘marking stick’ ‘pencil, pen’
c. erihtł’ís net’į ‘marking that’s looked at’ ‘movie, TV’
d. náke saritł’ésé ‘two sun-marking’ ‘it’s two o’clock’
3.5.5 Processual antonymy through negation
In §3.5.2, we considered the effects of suffixation on nominal stems to create new lexical items.
Here, I present some stative and process verbs with and without the all-purpose negativizing
suffix, -(h)íle. The examples in Table 4 are fully conventionalized in the language and do not
have the feel of a derivation, as I try to indicate with the English glosses.
Table 4. Neologisms based on the negativizing suffix, -(h)íle.
bare stem
with negativizer: -íle ‘not’
dánechá ‘they’re big’
dánechíle ‘they’re small’
ʔełtth’i ‘it’s right’
ʔełtth’íle ‘it’s wrong’
nedáth ‘it’s heavy’
nedádh’íle ‘it’s light’
benasní ‘I remember 3SG’
henasníle ‘I forget/I don’t remember’
bek’éghesní ‘I kept/took care of 3SG
bedóghesníle ‘I ignored 3SG’
bóreni ‘it is easy’
bóreníle ‘it is difficult’
dáhoní ‘they know it (the place)’
dáhoní híle ‘they’re absent-minded’
3.6 The temporal landscape and time expressions
It should come as no surprise that many temporal expressions in Dene Sųłiné are figurative since
the lexicalization of concepts in the time domain is one that is particularly susceptible to
metaphoric treatment. A small set of temporal expressions––mostly full propositions––involve a
variety of ontological TIME IS A SUBSTANCE/LOCATION metaphors and PROCESS FOR
AGENT/OBJECT/LOCATION metonymies. The first example below, in (96a), is a variant of the
stereotypical expression attributed to all indigenous North Americans for ‘several months ago’:
many moons ago. In the Dene Sųłiné example, the earth, rather than the moon, stands for a year,
rather than a month. The remaining expressions use spatial language to locate a place in time.
TIME IS SPACE
(96) a. įła néné ‘one earth’ ‘one year’
b. yanathé horeltth’ełi ‘that which will happen next’ ‘the future’

41
c. yánáth xáye, yánáth néné ‘next winter’, ‘next earth’ ‘next year’
d. dzinék’e ‘on the day’ ‘today’
e. tthi dzinék’e ‘ahead of/before today’ ‘yesterday’
3.6.1 The passing of time
Most expressions in Dene Sųłiné involving the passage of time (or the aging process in living
creatures) are based on a complex of verb stems, {-thi, -dhi, -ther, -dher}, associated very
generally with undergoing (in the absence of the pre-stem transitive or causative valency marker
-ł-) and doing (in its presence). Perhaps we could say that, at a very abstract level, these
associated stems predicate something about change of state, which can come about either
through motion (intransitive) or activity (usually transitive).17 All of the examples in (97) feature
some inflected form of this ‘undergo/do’ stem complex, while the specific lexical differences are
due to the individual prefixes or other incorporated material. Most important, for present
purposes, is that these expressions have to do with the passage of time. Again, in Dene Sųłiné as
in so many other languages, time is predicated against a spatial landscape: a
human/celestial/temporal subject moves and time passes. Rather than ‘undergo’ or ‘do’, I will
use the verb ‘move’ in all the literal glosses.
PASSAGE OF TIME IS DISLOCATION IN SPACE
(97) a. núnesthi (<níhónesthi) ‘ground-I start to move there’ ‘I’m aging’
b. nóníldher ‘ground-we (DL) moved there/fell down’ ‘we are old’
c. dzine hoghįdher ‘day-it has moved there’ ‘the day is over’
d. nahódher ‘AREAL had moved there’ ‘it (time) happened’
3.6.2 Times of the day
All of the following predications (all are full propositions) in (98) again share the same verb
stem, in this case either the imperfective -ʔa or its progressive counterpart -ʔał. This stem is a
common classificatory verb stem that signals that some entity construed as a solid round object
(SRO) moves or is moving. In all but one of these expressions, the sun––the presumed referent
for this classificatory verb stem––is not stated overtly. However, the motion of the sun in the sky
during the course of a day stands metonymically and metaphorically for the passage of time.
17 In his catalogue of Dene Sųłiné stems, Li (1933:139) wrote, “[t]his stem (sic) refers probably to some
general idea of activity, its meaning depends more upon its prefixes; thus, it may mean ‘to wake up’ (ø-
intr; ł-tr); ‘to die’ (ø), ‘to kill’ (ł); ‘to desire’, ‘a snow-storm comes’, ‘to snow’, ‘sickness comes’, ‘to be
sick’, ‘to suffer’, ‘to echo’, (ø); ‘to move, travel’, ‘to become old’, ‘to notify, to pass the news’, ‘to render
service’, ‘to struggle’ (ł).” To add a further level of complication, the stem forms for SG/DU subjects are
different from those for PL subjects. Such person- and number-based stem suppletion is quite common in
Athapaskan languages for certain verbs.

42
TIME PASSING IS THE MOVEMENT OF THE SUN (OR ITS METAPHORICAL EXTENSION, CLOCK)
(98) a. xaʔa ‘it (SRO) comes out’ ‘it’s dawn’
b. nidhaʔał ‘it (SRO) is far away’ ‘it’s noon’
c. ts’eʔał ‘it (SRO) is bright/clear’ ‘the sun is shining’
c. nuyeʔą ‘it (SRO) moves into the earth ‘it’s dusk’
d. nághįʔą ‘it (SRO) goes down’ ‘it’s evening’
e. sa gheʔał ‘the sun, it (SRO) is moving’ ‘the clock is ticking’
3.7 Miscellaneous states and processes
I conclude this lengthy section with a number of diverse expressions that involve the attribution
of some common processes––the expression of which is largely metaphoric.
3.7.1 The life cycle: Love and death, etc.
There are a number of figurative expressions pertaining to stages and milestones in a person’s
life, including the beginning and the very end, both of which transitions are conceived of as
movement to a place, as shown in (99) and (100):
CHANGE OF STATE IS CHANGE OF LOCATION
(99) a. xáilge ‘I crawled out’ ‘I was born’
b. nǫ(k’e níniya ‘I arrived on earth’ ‘I was born’
(100) a. łeghánįdher ‘3SG moved to the end’ ‘s/he died’
b. łeghániłther ‘I moved 3SG to the end’ ‘I killed him/her’
c. denedhíé ‘person’s movement/suffering?’ ‘death’
Next, we consider a variety of interpersonal situations: the loss of virginity (for males, at least!)
in (101a); the tricking or cheating of another person, as in (101b); or the uniting of two people as
husband and wife, as in (101c-d).
ONTOLOGICAL METAPHORS
(101) a. ts’a aze áyehésyél ‘he lost his little hat’ ‘he lost his virginity’**
b. hoya ghiłtį ‘I handled.ANIMATE into a hole’ ‘I tricked/deceived him/her ’
c. ełghąnihįt’as ‘they (DU) sat together’ ‘they got married’
d. beghą( nesdá ha ‘next to 3SG, I sit down FUT’ ‘I will marry him/her’
All of these processes are lexicalized via ontological metaphors. In the first case in (101a), the
loss of something intangible is being made tangible. In the second case in (101b), manipulating
someone through deceit means having power over the person in an one-upmanship sort of way,
much as if the person were placed in a hole, a “one-down” position, to be sure. In the third case,
in (101c-d), sitting with or sitting next to someone (a physical act) stands for getting married (an
abstract interaction). Perhaps this case might be better described as an elaborate phase metonymy

43
on par with the English expression, walking down the aisle, which merely describes an initial
episode to the whole event of getting married (and getting married, specifically, in a church).
3.7.2 A day in the life
Many everyday activities also have a figurative cast to them. I would argue that metonymy is the
dominant semantic mechanism at play in these expressions, although some have a slight
metaphorical feel to them (e.g., ‘moving into the clear’ for ‘waking up’). For the most part, these
actions are metonymic because the verb (and in some cases, the subject) actually lexicalized is
only indirectly associated with the conventionalized meaning. More than GENERIC FOR SPECIFIC
or SPECIFIC FOR GENERIC metonymies, the examples below tend to lexicalize an initial or
intermediate phase of the process, leaving the resulting state to be inferred.
WHOLE FOR PART (CONSCIOUSNESS) and PART (HEAD) FOR WHOLE
(102) a. ts’ehedhi ‘(in) clear-3SG moves’ ‘s/he is waking up’
b. ni-tthí-rést’a ‘up-head-I.moved.SRO’ ‘I got up (from sleeping)’
c. daghéghesʔá ‘I positioned.SRO (head) upward’ ‘I laid on my back’
ASSOCIATED PROCESS FOR PROCESS (LYING DOWN IS SLEEPING; SLEEPING IS DREAMING)
(103) a. thetį ‘3SG is lying (down)’ ‘s/he is sleeping’
b. náthetį ‘3SG slept twice’ ‘s/he dreamed’
ASSOCIATED PROCESS FOR PROCESS (DRINKING IS DRINKING TO EXCESS; DRINKING IS DROWNING)
(104) a. yenéhdą ‘3SG drank of it’ ‘s/he is drunk’
b. tuhedą ‘3SG drank water’ ‘s/he drowned’
INITIAL PHASE FOR ENTIRE PROCESS (GETTING DRESSED)
(105) a. nástł'u ‘I’m tying up’ ‘I’m getting dressed’
b. yu yessa ‘clothes, I’m going into’ ‘I’m getting dressed up’
c. ʔih náréghesyá ‘coat, I went up/handled into it’ ‘I wore a coat’
d. t’así náréghesʔą( ‘UNSPEC I handled.SRO up (on head)’ ‘I wore s.th. on my head’
e. ke yéghesʔédh ‘shoes, I kicked into (them)’ ‘I wore shoes’
f. thedh hetheréłtł’ǫ ‘belt, 3SG tied it on’ ‘s/he wore a belt’
g. beyédesnígh ‘I manipulated into it’ ‘I wore it (gloves/mitts)’
3.7.3 Action and interaction
Expressions about what people do and how they interact with each other are often formulated
using figurative language. Related to the psychological state predications explored in previous
subsections are a host of expressions relating to physical processes or states. Some of these allow
one physical process to stand for another, more abstract or habitual one, as in the case of a
qualitative assessment, for example. Others involve attitudes or more mental events. They all are
expressly figurative by virtue of the many metaphors and/or metonymies that underlie them.

44
Lexicalized expressions with the complex verb stem yati ‘talk’ are especially numerous, as
shown in (106). All of these expressions are built on a set of basic metonymies by which the
speech process stands for the patient (106a), the agent (106b), or the manner (106c) of speaking
(based on the content of the talk). The example in (107) is a nice instance of an ontological
metaphor whereby language is construed as a physical object. This is a common metaphor across
the different Athapaskan languages––conceiving of language as something tangible which can be