Navigation – Plan du site

AccueilNumérosVolume 2Conceptual transfer and the emerg...

Conceptual transfer and the emergence of the sign

A semiogenetic approach to PIE *ĝenu- ‘jawbone, chin’ and *ĝenu- ‘knee’
Dennis Philps


Dans le cadre d’une théorie sémiogénétique des conditions d’émergence et d’évolution du signe, nous proposons que l’invariance phonologique caractérisant les racines proto-indo-europeénnes *ĝenu- ‘mâchoire, menton’ et *ĝenu- ‘genou’ est la trace en surface d’une invariance notionnelle sous-jacente. Cette proposition intègre, outre ses dimensions linguistique et sublinguistique, plusieurs dimensions extralinguistiques : neuroscience, cognition, mythologie, etc. Nous soutenons enfin que l’hypothèse de l’existence d’une stratégie de nomination inconsciente du corps, fondée tant sur une représentation somatotopique de celui-ci que sur des principes d’auto-référentialité et de transfert conceptuel, constitue une explication plausible de l’identité de forme attestée par ces deux racines, mais aussi des liens qu’elles pourraient entretenir avec *ĝenh1- ‘engendrer, naître’, *ĝneh3- ‘connaître, reconnaître’ et *gwénha- ‘femme’.

Haut de page

Notes de la rédaction

Soumis le 06/12/2006
Version révisée reçue le 17/10/2007
Version finale 11/06/2008

Avertissement : cet article contient des caractères spéciaux au format Unicode. Si vous rencontrez des problèmes d’affichage, vous devez probablement télécharger et installer une police Unicode ; vous trouverez une police Unicode gratuite et libre de droits à l’adresse suivante :

Disclaimer: this article contains special characters encoded in the Unicode format. If you have trouble reading this text, you should probably download and install a Unicode font; you can find a free software Unicode font at the following URL:

Texte intégral

1. Introduction

1This study of the possible relations between the Proto-Indo-European roots *ĝenu- ‘jawbone, chin’ and *ĝenu- ‘knee’, and of the role that the cognitive process of conceptual transfer may have played in establishing these relations will be conducted within the semiogenetic theory of the conditions of emergence and evolution of the linguistic sign (STEELS) sketched out in Philps 2006. In reference works such as Pokorny 1959 and Watkins 2000, these roots, considered as homonymic, are listed separately. The main claim I shall be making here is that their identity of form is the surface trace of a deep-level notional invariance based on conceptual transfer patterns within the human body (Heine 1997: 133).

2Several scholars have already suggested that the roots in question are linked by a common notion. Referring to PIE *ĝenu- ‘jawbone, chin’, Buck, for instance, proposes that an “Orig. sense ‘jaw’ and ultimate connection with the word for ‘knee’, IE *ĝenu- (neut.), Grk. γόνυ, Lat. genū, etc. […] through some common notion like ‘angle’, is probable.” (1949: 221). This proposal is also put forward by Mallory & Adams: “*ĝónu- (gen. *ĝénus) ‘knee’. […] Most probably related to *ĝenu- ‘chin, jaw’, both being sharply angled parts of the body. […] Words for the ‘knee’ were often used euphemistically for the genitals.” (1997: 336). Moreover, Ernout & Meillet suggest in their Dictionnaire étymologique de la langue latine, under genae, genārum ‘cheeks’ (f. pl.), that the form genu- in genuīnī dentēs is

“doubtless from the same family as γωνία ‘angle’, like genū [‘knee’]. Lat. gena may be explained by its female gender ; cf. nurus, nora ; its form enabled a distinction to be made between the name for ‘cheek’ and that for ‘knee’” (2001: 270, my translation)

3I will begin by surveying the mythological and morphological dimensions of the problem, before presenting the semiogenetic hypotheses that provide the theoretical framework for this paper, particularly the notions of self-referentiality (Searle 1983), conceptual transfer (Heine 1997), vocomimesis (Donald 2001), and ‘sublexical marker’ (Philps 2006).

2. The mythological dimension

4In the early part of the 20th century, certain scholars put forward the hypothesis that *ĝenu- ‘knee’ is cognate with *ĝenh1- ‘to beget a child, be born’ or even *ĝneh3- ‘to know, recognize’. For instance, Loth points out that in Irish, glúinighim (v.), derived from glún ‘knee’, means ‘I beget’, and that in other Indo-European and non Indo-European languages, there is

“converging evidence that proves that the two meanings of the word genou [‘knee’] had a very different origin in prehistoric times […] In those days, a woman would give birth while sitting on her husband’s knees, the husband thereby acknowledging that the child was indeed of his own blood and that it would become part of his family. Subsequently, the mere act of the father sitting the child on his lap came to have the same meaning.” (1923: 145, my translation)

5In a similar vein, Benveniste states that Old Irish glún-daltae ‘nursling born from the knee’ and Sogdian z’nwk’ z’tk ‘son born from the knee’ mirror each other: “Like Old Irish in Celtic, Sogdian is, in Iranian, the only dialect to attest a comparable use of the name for the ‘knee’.” (1927: 53), while according to Cahan, “the human body is what provided the Germanic tribes with all the images that serve to represent the structure of the family and the organization of kinship.” (1927: 66). Meillet puts forward the idea that a child described as genuīnus “must originally have been one who, having been sat on the lap of the family’s father, was henceforth acknowledged as being his descendant.” (1927: 54, my translation). The author continues

“The explanation which consists in justifying a possible relation between the group formed around genu and that formed around γίγνομαι by a practice of childbirth […] does not seem particularly convincing. […] Consequently, the group formed around Lat. genu is best seen as related to the root *g’enə- *g’nē- ‘to know, recognize’, rather than to *g’enə- *g’nē- ‘to beget’. This being posited, one is led to wonder whether the two roots *g’enə- ‘to know, to recognize’ and *g’enə- ‘to be born, to beget’, absolutely identical in form, but different in the historical period, were not originally one and the same.” (1927: 54-55)

6Lastly, Meunier states that words deriving from the PIE root he notes as *gan (‘to give birth, beget’) include Classical *έ-γει-νά̊μην, Lat. gen-uī ‘procreate’ and Eng. queen (1871: 73), although Mallory & Adams (1997: 648) derive the latter from *gwénha- ‘woman’, a form which they prefer to treat as unanalysable. Be that as it may, one cannot dismiss out of hand, in the light of the above, the possibility that *gwénha- was related, at some indeterminably early stage of language prior to Proto-Indo-European, to *ĝenu- ‘jawbone, chin’, *ĝenu- ‘knee’, *ĝenh1- ‘to beget a child, be born’, and *ĝneh3- ‘to know, recognize’. Even though such a link is not linguistically demonstrable, it is worth noting that the C1 slot in all these roots is occupied by a voiced tectal occlusive (an older term for ‘plosive’), whether labialised (*gw-), aspirated (*gh-), or not (*g- in later forms). In this respect, Buck’s comment that *gwénha- is “[w]ithout known root connection, unless at some remote period, before differentiation of the gutturals (?), with IE *ĝen- ‘beget’” (1949: 82) may well be pertinent, as may Koch’s speculation that

“The vocal signifier /kuna/, /kwena/ [> Gk. γυνή, Goth. qinō, Eng. queen, etc.)] thus came to be associated with such meanings as ‘bending’, ‘knee’ etc. What makes the use of bending knees prototypical for ‘woman’ is the process of ‘giving birth to babies’. Even modern European women, when in labour, naturally press their hands and bend and press their knees; but they lie on their backs when they do so. In earlier times and even today, in more primitive societies, women resort to a more efficient position of birth-giving: they stand semi-erect or squat.” (2001: 24, see also 2008).

7Whatever the hypothesis favoured, it seems clear that, as noted by Onians (1988, chap. 2), the human body underwent a complex process of mythification in prehistory, anchored in the knowledge and beliefs of the periods concerned. Onians, however, is taxed by Chantraine in his Dictionnaire étymologique de la langue grecque with proffering certain etymologically rash considerations (1999: 233), and Chantraine himself does not actually posit a connection between γνάθος ‘jaw’ and γόνυ ‘knee’. What exactly are these considerations? Onians first quotes Aristotle, for whom “boys maimed in their sexual organs are sterile (ἄγονοι) so that they do not get a beard (γενειαν) but continue as eunuchs” (De Gener. Anim. 746 b, 21 ff). Based on this quote, and on the idea that the synovial liquid in the knee, whose real function was unknown at the time, was classed with the “cerebro-spinal fluid or liquefiable substance, the fluid of life, and thought to contribute to the seed.” (1988: 177-178), he goes on to propose the following explanation for the connection between the jaw, the knees, and generation in prehistoric times

“This evidence suggests what does not appear to be recognised, that the jaw like the knee was associated with generation and that the other names of the chin or jaw, γένυς and γένειον (cf. γενειάς, ‘beard’, Latin gena, German Kinn = chin, A.S. cinn, etc.), were related to γένος, γενεά, genus, Kind, ‘kin’, etc., expressive of generation.” (1988: 233)

8Although this hypothesis is etymologically shaky, it does rest on an in-depth study of the works that mark the beginnings of European civilization, such as Homer’s Iliad and his Odyssey. It also calls to mind the fact that in English, according to The Oxford English Dictionary (2nd ed., 1989, henceforth OED), the word marrowbone (usually pl.) also denotes, in jocular parlance, the ‘knees’, a sense first attested in 1532.

9Lastly, a link between the ‘knee’ and benediction has been observed in Semitic, as Hodge points out

“Semitic *b-r-k ‘bless’ has often been connected with *b-r-k ‘knee’ and *r-k-b ‘knee’ […]. This is erroneous. There are a number of Lislakh bases in **l-k. One means ‘angular, bent’ and is used for body parts involving an angle, such as ‘jaw’, ‘arm’, ‘hip’, and leg’. Semitic *b-r-k and *r-k-b are derived from this base” (1997: 209)

10For Bohas, certain words of Arabic with meanings referring to the region of the mouth or the knees belong to one and the same ‘matrix’ (μ), namely

{[labial], [dorsal]

the notional invariant of which is ‘curvature’. The author goes on to affirm that the ‘mimophonic’ property that he attributes to this matrix results from

11“the coupling of two articulatory properties, one external, the other internal. The former is the visibly curved position of the muscles of the mouth when phonemes displaying the feature [labial] are articulated, while the latter is the form taken by the tongue itself when a dorsal such as k or q is articulated.” (2000: 91, my translation)

12Among these words are afakku and fanîkun, both meaning ‘the place where the upper and lower jaw come together’ (semantic extension: ‘to open the hand, the mouth’ (forming a or a shape), and rukba ‘knee’ (rounded body parts).

3. The morphological dimension

13Towards the end of the 19th century, a few scholars began to study morphological data related to the names of certain body parts in the Indo-European languages in the light of the concepts of analogy and duality, although no attempt appears to have been made to theorize such data at the time. For instance, M. Bloomfield states that

“Designations of parts of the body exercise strong analogical influence upon one another, and occasionally the suffix of some one of them succeeds in adapting itself so as to be felt the characteristic element which bestows upon the word its value.” (1891: 3)

14The list of names of certain body parts in the classical languages attesting an ancient *r/*n heteroclitic declension which is given by Bloomfield includes words denoting the mouth, the jaw(s), the mouth, and the cheek(s), among others. The author further points out that the n- stems appearing in the oblique cases of this declension are also to be found in the paradigms of other heteroclitic declensional types in body-part words. Significantly, the list he provides in support of his claim once again includes words in the classical languages some, though not all, of which denote protrusive, articulated, bony parts of the body, whether rounded or angular, e.g. ‘ankle’, ‘breast-bone’, ‘collar-bone’, etc. It should be remembered in this respect that Watkins (2000: 26) reconstructs the original meaning of the PIE root *ĝenu-, with derivatives referring to the chin, the cheeks and the mouth, as ‘jawbone’ (and not simply ‘jaw’). Also, Cuny affirms that “[T]he names of parts of the body which come in pairs provided a pattern for the subsequent formation (and the ‘description’) of other body-part words.” (1924: 181, italics in original, my translation). He backs up this remark with examples such as Gothic kinnus, Latin *gen-u-a (cf. gen-a), whence genuīnī (dentes) ‘one of the two jaws’, Old Indic jā́nu, Gothic kniu (*gnew-o-m), etc... ‘one of the two knees’, all attesting -u- (-ū-), which the author motivates by linking it to PIE *dwo ‘two’ (1924: 26).

15The above data and the analogically or quantitatively based processes they attest result from morphological processes undergone by the body-part nouns in question. It may be noted that each of the paired body-parts in question is characterized, relative to other parts of the body, by a number of invariant anatomical properties that are either general (articulation, angularity, protrusion, symmetry, movement, etc.) or specific (location, function, etc.). As will be seen, the semiogenetic theory constraining this study posits, notably on the basis of homologies characterizing various parts of the human body, the bodies of animals, and non-animates denoted by one and the same word in some Indo-European languages, e.g. Eng. head (of a human being, of an animal, of a nail), that the human body may have been used as a structural template not only for self-naming purposes (Heine 1997: 137), but also to ‘name’ animals and non-animates.

16One scholar to have evoked this possibility, though once again without theorizing it, is Cuny, who suggests that

“One can also surmise that in the minds of children, the names of body-parts (and even tools) were assimilated to those of animals, a given part of the body itself being homologized with that of an animal, a tool with a living creature, etc.” (1924: 181, italics in original, my translation)

17Referring to the names of tools in the classical languages that also denote certain parts of the human body, Cuny further asserts that

“Moreover, one cannot keep these words apart from the name for the ‘knee’: Old Ind. jā́nu, Gk. γόνυ, Lat. genu, Germ. *kn-ew-an (Goth. kniu, etc.). Here again, the same close connection between a tool name and a body-part name is to be found, since in Greek, γωνία […] is related to γόνυ, hence furnishing the noteworthy semantic proportion set out below:
γωνία ‘angle’ is to Lith. z̆am̃bas ‘ridge (of a beam)’
γόνυ ‘knee’ is to Lett. ∫ůb-s, Old Sl. zǫ ‘tooth’” (1924: 159)

4. The phylogenesis of language

18The connection between tool names and body-part names noted by Cuny is reflected to some extent in theories of the phylogenesis of language, including those propounded by Paget (1930), Jóhanesson (1949), Corballis (2003), and Philps (2006), particularly as concerns the relation between articulatory gestures on the one hand, and manual gestures, particularly those that involve crafting (and hence the tools required for this activity), on the other. Paget, for instance, states that

“As man developed in intelligence and technique, the need for more exact gestures became greater, while his hands became more occupied with the arts and crafts. Gestures, which were previously made by hand, were unconsciously copied by movements or positions of the mouth, tongue or lips”. (1930: 132),

19while Jóhanesson asserts that the speech organs tend to move in unison with hand movements

“When primitive man first learned to form sounds, it was quite natural that the speaking organs should imitate the method he used before learning to speak, the movements of the body and especially the gesticulations of the hands” (1949: 81)

20Philps, whose hypothesis takes account of results obtained by Gentilucci et al. (2001) in laboratory experiments, has postulated that if open-close hand gestures were transformed into open-close mouth gestures, as Gentilucci’s conclusions suggest, then the occlusive sounds produced by these articulatory gestures could, since they are synchronized, have served to refer back to the hands and the gestures they perform deictically, and to stand for them iconically by means of an unconscious, neurophysiologically-grounded body-naming strategy (2006: 248). This strategy may subsequently have been extended, by conceptual transfer, to tools and other objects with which the hands interact, leading to the resultant linguistic signs being used to denote both the hands (or hand gestures) and the entity grasped, with or without overt morphological marking (cf. Mod. Eng. hand / handle (n.) and grasp (n.) ‘the grip of the hand’ / ‘that which is fitted to grasp […], a handle’, OED).

21To return to PIE *ĝenu- ‘jawbone, chin’ and *ĝenu- ‘knee’, it may be noted that, construed in this manner, certain derivatives of the latter bear witness to a cognitive process of ‘body bleaching’ (or decorporealization) as their denotative function is extended from body-internal to body-external referents. One such derivative is Eng. knee ‘A joint in an animal likened to, or regarded as corresponding in position or shape to, the human knee’, ‘Part of a hill, tree, etc., regarded as corresponding to the knee’, and ‘A piece of timber having a natural angular bend, or artificially so bent’, i.e. an object whose form and function resemble those of the knee (OED). As for *ĝenu- ‘jawbone, chin’, certain derivatives of Gk. γένυς ‘jaw(s)’ also attest this phenomenon, e.g. γενειαστήρ ‘chin strap’ (part of a (double) bridle), the suffix of which was used to provide tool names, according to Chantraine (1999: 215).

5. The concept of ‘sublexical marker’

22The semiogenetic hypotheses which provide the theoretical framework for this paper stem from an analysis of those word-initial consonant segments of English with recurrent form and meaning called ‘phonæsthemes’ (e.g. br-, gl-, sl-, sn-) since Firth coined the term (1930), although these ‘frequently recurring sound-meaning pairings’ (Bergen 2004: 290) were identified as long ago as the middle of the 17th century (Wallis 1653). In view of the lack of any rigorous definition of phonæsthemes or of the criteria for classifying words containing them, I applied a principle of invariance to the words in which they are found. In mathematics, an invariant property or quantity may be defined very generally as one that remains unaltered by a given transformation (or mapping), especially a linear one, while in cognitive linguistics, Lakoff’s Invariance Hypothesis (1990) posits that cross-domain mappings preserve the cognitive topology of the source domain.

23Stated briefly, this procedure allows one to identify certain subsets of words attesting a given word-initial phonæstheme (e.g. sn- in the heuristic class of ‘sn- words’) whose members attest both phonological invariance (e.g. /sn/) and notional invariance (e.g. /nasality/ in the subset of ‘sn-words’ composed of sneeze, sniff, snore, etc.). A ‘notion’ is envisaged here as a cognitive entity which may be defined as a set of complex mental representations resulting from the embodied mind’s attempts, necessarily conditioned by human subjectivity, culture and convention, to conceptualize and categorize experience, notably the formal, functional, compositional and spatio-temporal aspects of the latter. Seen in this light, notions may be analysed into ‘notional domains’, noted between slashes (e.g. /nasality/), to which a metric may be applied (Gärdenfors 2000: 17, 26).

24I refer to the word-initial consonant segments thus conceptualized as ‘sublexical markers’, defined as submorphemic units displaying phonological and notional invariance within the subset(s) of words in which they occur. Sublexical markers are noted herein between angled brackets, e.g. <sn->. Although <sn-> happens to correspond to the word-initial phonological structure /sn/, this is not always so, since the marker <gn->, for instance, which occurs in a small subset of ‘gn- words’ related to /biting/, e.g. gnash, gnat (‘a biting dipteran insect…’), and gnaw, no longer corresponds to the word-initial phonological structure /gn/ as it did in earlier stages of the language (Dobson 1968: 977-978), but to /n/. In this case, as in that of kn- in ‘kn- words’, it is necessary to mobilize the concept of ‘semiological invariance’ (gn-) for purposes of diachronic analysis.

25Now in a small number of English ‘sn-/n- doublets’, e.g. sniff/niff, sn- appears to function as a ‘core invariant’ and s as an initial variant, since the presence or absence of /s/ does not change the underlying notional domain common to each member of the doublet (in the above example, /nasality/). A core invariant may be defined as the smallest element in the phonologically invariant, word-initial consonant segment of a given class of words (e.g. n in ‘sn- words’) to which a common notion (e.g. /nasality/) may be attributed. The ‘sn-/n- relation can therefore be hypothesized as being subsumed at marker level by an <sn->/<øn-> relation, where <ø> represents a slot for possible consonant alternation which remains uninstantiated in neeze, schematically øneeze.

26Diachronically speaking, the submorphemic relation between <sn->/<øn-> and /nasality/ (noted {(s)n-, /nasality/}) can be shown to go back, via Proto-Germanic, to Indo-European (Southern 1999: 70-71, *(s)neu̯- and its various extensions), as can that between <gn-> and /biting/ (cf. gnathic ‘of or pertaining to the jaws’), derived, via Greek, from *ĝnḥa--dh-o-s ‘jaw’). However, due to the diverging evolution of the two complementary base forms (I: CVC-, II: CCV-), the relation between <gn-> and /knees/ attested in PIE (cf. *ĝnéu-, a variant form of *ĝenu- ‘knee’) has become, in English, a relation involving both <gVn-> (cf. genuflect (v.) ‘to bend the knee…’ < med. Lat. genūflectĕre < Lat. genū ‘knee’ + flectĕre ‘to bend’) and, as a result of the devoicing of root-initial *g- between PIE and Proto-Germanic, <kn-> (knee < Proto-Germ. *knewam < *ĝenu- ‘knee’). This should not be construed as implying that all the English ‘sn- words’ or ‘gn- words’ concerned can be reconstructed for PIE or even for Proto-Germanic, since it is highly likely that many are late, analogically-based innovations.

6. STEELS and self-referentiality

27Assuming that the {(s)n-, /nasality/} relation can indeed be reconstructed for Indo-European, and that it may therefore be incommensurably older, I postulate that the phonetic realization of the core invariant in <sn->, i.e. <n>, is endowed with a self-referential potential. In other words, given that this realization corresponds to a sound produced by air passing through the nasal cavity, the soft palate being in its lowered position, and that this has always been so, phylogenetically speaking, it may have served as a natural ‘building-block’ around which words conveying nasal senses could be constructed. This is not to imply that such a possibility is necessarily materialized in any given language, or that any occurrence of /n/ in a given system, irrespective of its position, necessarily attests this potential. In Indo-European, for instance, *-n- was root-infixed to form a present from the verb stem (Szemerényi 1999: 270) independently of the notion conveyed by the root.

28Furthermore, there is structural evidence in PIE, notably root-final *-r-/*-l- alternation that does not correlate with a change in ‘meaning’, as in *gal- ‘to call, shout’ / *gar- ‘to call, cry’ (Benveniste 1964: 25), that in certain cases, the consonant which occupies the C1 slot in the canonical PIE root structure C1eC2- (e.g. *ĝ- in *ĝenu-), functions as a ‘core invariant’ (here <ĝ>), and *-r-/*-l-, consequently, as variables (C2). I therefore postulate that the phonetic realization of the core invariant <ĝ> in *ĝenu- is endowed, like <n> in <sn->, with a self-referential potential. By reanalysing this root-initial occlusive as a gesture of occlusion (Browman & Goldstein 1992: 156), and hypothesizing that it originally served to refer back to the goal-orientated, open-close mandibular movements involved not only in non-linguistic oral activities such as biting, but also in pre-linguistic oral activities such as primitive calling, we may trace it back to a possible scenario of the emergence of human speech similar to that described by MacNeilage in his “Frame-Content Theory” (1998, see also MacNeilage 2008).

29It is on this principle of self-referentiality (see Searle 1983: 222-224) that my semiogenetic theory (STEELS) is founded. The theory, constrained by the limits imposed by the Indo-European corpus, posits that the sign was originally configured vocomimetically (Donald 2001: 291), during an indeterminably early period in the evolution of language when the vocal apparatus was used for nutrition, respiration, and visuofacial communication. As Ladefoged points out

“None of the vocal organs was meant for talking. The lips are there to keep the mouth sealed off when we are not eating. The tongue and the teeth are for making food more digestible. The prime purpose of the vocal folds is to stop foreign matter entering the lungs. And the lungs are for breathing, not for pushing air out to make noises.” (2001: 113)

30The very essence of this principle requires us to take account of the individual phonetic features characterizing each core invariant in order to identify its possible self-referential potential. However, this is clearly insufficient, not least because in phonology, distinctive features are static, resultative entities representing final states of articulation, e.g. [anterior], [coronal], [high], and their combinations.

31Now if it is posited that the sign originated vocomimetically, then it becomes necessary to take into account the very nature of mimicry in Homo Sapiens. Mimicry is defined as ‘The action, practice, or art of mimicking or closely imitating […] the manner, gesture, speech, or mode of action of persons…’, and imitation as ‘The action or practice of imitating or copying’ (OED). Since mimicry and imitation are actions, it also appears necessary to take account not only of the phonetic features of core invariants, but also of their kinesic aspects. If we examine the status of the latter in the light of Joly & O’Kelly’s psychomechanical approach (1990: 28-30), we are led to the conclusion that for a given phonetic feature to exist, the conditions leading to its existence must have already been present, since the condition necessarily precedes the consequence. In particular, the feature [occlusive] can exist if and only if the vocal apparatus has previously performed an occlusion at some point along the vocal tract. Now an occlusion is not only an articulatory gesture, but also a masticatory gesture, mastication being defined as “the action of […] chewing”, and the act of chewing as “To crush, bruise, and grind to pulp, by the continued action of the molar teeth, with help of the tongue, cheeks and saliva” (OED). This dual role is highlighted in the entry given by Trask (1996: 246) : “Occlusion n. 1 A complete closure within the mouth during an articulation, as occurs during a plosive. 2. Complete closure of the jaws, bringing the teeth together.”

32Within STEELS, the postulate that the earliest, prelinguistic vocalizations of Homo Sapiens were self-mimetic is based on the observation that the articulators forming the human vocal tract are the only bodily organs that are potentially self-referential in terms of the production of sound. Hence the dental sound produced when the tip (or blade) and rims of the tongue are raised from their position of rest to make contact with the upper teeth ridge and side teeth can be projected metonymically, i.e. the effect for the cause (here the sound articulated for the articulators), onto the active or passive articulators (here, the tongue and the teeth respectively). This may have resulted in the sound thus produced functioning as the core of a self-referential sign allowing the speaker to ‘name’, and the hearer to identify, by intentional attunement (see Gallese 2005), not only the articulators involved, but also, again by metonymy (spatial contiguity), the various properties of the notional domains associated with the production of this type of sound.

  • 1 See Martinet 1957.

33Moreover, returning to the notional domain /biting/, it so happens that all those PIE roots whose derivatives relate to this domain attest an occlusive sound in word-initial position, whether a plosive (e.g. *denǩ- ‘to bite’, Rix et al. 2001: 117-118), a resonant (e.g. *mendh- ‘to chew’, Watkins 2000: 55), or, possibly, a laryngeal (cf. Lehmann 1993: 86). With reference to the ‘naming’ process posited, it may be noted that in *(h1)dont- ‘tooth’ and *dn̥ĝhū- ‘tongue’, the consonants are all linguo-dental, with the exception of the laryngeal in *(h1)dont-, the phonological status of which is not clear,1 and *-ĝh- in *dn̥ĝhū-, whose internal structure has been analysed notably by Hilmarsson (1982) and Martinet (1986). With respect to *dn̥ĝhū-, Martinet asserts that “There is every chance that this root was originally a compound.” (1986: 128, my translation). These linguo-dental consonants are also coronal, since the tip (or blade) of the tongue is raised towards the upper teeth ridge, thus initiating an articulatory gesture of protrusion within the oral cavity. Likewise, the occlusive sounds produced by a complete closure in the vocal tract followed by the release of accumulated air pressure may have come to function self-referentially as a natural ‘building-block’ around which words with occlusive senses could be constructed.

  • 2 See Bear, Connors, & Paradiso 1996: 413-415.

34To take PIE *ĝenu- ‘jawbone, chin’ and *ĝenu- ‘knee’ as examples, it may be hypothesized that when realizing their putative core invariant <ĝ> semiogenetically, the vocal organs were unconsciously simulating, by means of a cyclical, articulatory movement of occlusion involving the jaws and the dorsum (i.e. constriction-release), the cyclical, articular movements of the knees (i.e. flexion-extension). It may further be hypothesized that this process is somatotopically grounded, that is to say, based on a topology-preserving mapping of sense organs on the somatosensory cortex.2 Words formed in this way would be expected to have undergone conventionalization and phonetic modification as system-internal conditions gradually obscured the original, system-external conditions, as when the initial dental in PIE *dn̥ĝhū- ‘tongue’ became a lateral approximant in Latin lingua via a putative form dingua, perhaps under the influence of Sabine dialects and lingere ‘to lick’ (Hilmarsson 1982: 356).

35Given that the postulated naming strategy rests on principles of self-referentiality and metonymy, and insofar as the latter is based on relations of association and contiguity, it may be hypothesized that this strategy was extended, for ‘naming’ purposes, from the interior to the exterior of the vocal organs. Indeed, the relation between interior and exterior underpins every speech act, since the product of the latter is determined, physiologically speaking, by the internal and external characteristics of the different resonators. This metonymic extension may be illustrated by the meanings acquired by various derivatives of PIE *ĝenu- ‘jawbone’, but also ‘chin’, e.g. ‘cheek’ (Latin gena) and ‘mouth’ (Old Irish gin).

36An approach that parallels STEELS to some extent is MacNeilage’s aforementioned “Frame/Content Theory”, which holds that the mandibular cycle is the main articulatory building-block of speech, and proposes that “[syllabic] frames may derive from cycles of mandibular oscillation present in humans from babbling onset, which are responsible for the open-close alternation.” (1998: 499). By this view, human speech seen as a motor function emerged when the cyclical, open-close alternations characteristic of ingestive activities underwent a series of adaptations resulting in them being employed in content-modulated syllabic frames for purposes of visuofacial and phonatory communication.

7. Conceptual transfer

  • 3 See Morris 2002: 364-370 for the notion of body self-mimicry.

37STEELS posits that the naming strategy described above was generalized by a process of conceptual transfer to the human body as a whole. This process, physified by the vocal organs, can best be explained, in my view, by hypothesizing the existence of a capacity for self-mimetic simulation within the brain,3 the implementation of which would result in the holistic reproduction of the formal, functional and spatio-temporal conditions of operation of body parts in action, relative to the body as a whole. As Heine suggests, basic body-parts

“may serve as structural templates to denote other body-parts, as well as other items not connected with the human body—that is, concepts which are perceived to be related to the former with reference to shape, location, and/or function.” (1997: 134)

38Such a hypothesis appears to be accredited by recent research in the area of neuroscience. Feldman & Narayanan, for instance, state that

“There is considerable biological support for the idea of simulation semantics. The monkey studies discussed above have been interpreted as suggesting that the various action loops enable the animal to simulate and thus predict consequences of possible actions. A number of recent studies have suggested strongly that homologous circuits in human brains support similar abilities.” (2004: 390)

39The studies the authors refer to include those by Fadiga, Gentilucci, and Rizzolatti and their co-workers, on mirror neurons. With respect to the latter, Rizzolatti and Craighero state that

“It is plausible that, originally, the understanding of the words related to mouth actions occurred through activation of audio-visual mirror neurons related to ingestive behaviour” (2004: 185, see also Rizzolatti & Sinigaglia 2008)

40Within simulationist cognition theory, it has been proposed by Gallese (2005) that mirror neuron systems constitute the neural underpinnings of embodied simulation, and by Gallese & Lakoff (2005), that all conceptual knowledge is embodied, and hence that all cognitive operations are underpinned by sensory-motor representations.

41One of the predictions made by the above naming hypothesis is that, given the physiological configuration of the human body, some bodily properties would appear cognitively more salient than others. For instance, given the large number of bodily regions that are articulated and stand in some symmetrical relation to one another, e.g. the jaw(s), the knee(s), the ankle(s) and the elbow(s), it is to be expected that invariant formal and topological properties such as protrusion and angularity, as well as invariant functional properties (kinetic, prehensile, etc.), would be particularly salient, and that this would result in some form of linguistic salience.

42In a study of the major tendencies underpinning semantic change, Wilkins (1996: 274), echoing Matisoff (1978: 176-179), asserts that there are four categories of change, which are not mutually exclusive


intrafield metaphorical changes (i.e. within a given semantic field, such as the human body);


interfield metaphorical changes (i.e. between two different semantic fields, such as the human body and the body of an animal);


intrafield metonymic changes (i.e. within a given semantic field);


interfield metonymic changes (i.e. between two different semantic fields).

43It should be noted that whereas STEELS incorporates the concepts of ‘transfer’ and ‘notional domain’, Wilkins analyses ‘changes’ involving ‘semantic fields’. Furthermore, it seems that any metaphorical projection from a source domain to a target domain (Lakoff 1987) must be accompanied by concomitant metonymic projection, if one takes ‘metonymy’ to mean the internal relations, of whatever nature, that characterize the source domain and the target domain. Even if the metaphorical mapping theory set out in Lakoff (1987, 1990), Johnson (1987), and Lakoff & Johnson (2003) has been complemented by other theories, e.g. Gärdenfor’s conceptual spaces (2000), Fauconnier & Turner’s mental spaces (2002), and Lakoff’s own “Neural Theory of Language” (2003), its founding principles remain relevant, as does Heine’s theory of conceptual transfer.

44Lastly, with respect to *ĝenu- ‘jawbone, chin’ and *ĝenu- ‘knee’, the invariant properties of the two domains in question, i.e. the jaws, seen as the source domain of the transfer because of their potential to initiate the production of self-referential sounds in the vocal tract kinetically, and the knees, seen as the target domain to be named, would include topological ones such as protrusion and angularity, kinetic ones such as applied forces, and mechanical ones such as the degrees of liberty of the movements performed by these articulated regions.

8. Conclusions

45Within STEELS, the fact that PIE *ĝenu- ‘jawbone, chin’ and *ĝenu- ‘knee’ have been reconstructed as identical in form is one of the predictable consequences of the self-mimetic mapping strategy posited as having been implemented by Homo Sapiens since the earliest stages of its evolution, insofar as modern neurocognition allows us to advance this postulate. According to Damasio, for instance, “what ends up being mapped in the sensory regions of the brain and what emerges in the mind, in the form of an idea, corresponds to some structure of the body, in a particular state and set of circumstances” (2003: 197).

46A semiogenetic analysis of the available data suggests that *ĝenu- ‘jawbone, chin’ was built morphogenetically to what I identify as its core invariant (<ĝ>), by processes of syllabification (*ĝ- > *ĝV-, see Hurford 2003: 53), consonant incrementation (*ĝV- > *ĝVn-, see Southern 1999: 152), and, possibly, suffixation (*ĝVn- > *ĝVn-u-, see Cuny 1924: 26). In the phonetic realization of the core invariant <ĝ>, the occlusive (open-close) sound produced by the open-close movement of the jaws would have served to refer back to the jaws and their physiological environment (chin, cheeks…) iconically (Peirce 1991: 251-252). By this view, the ‘jaw(s)’ would have been named both self-referentially and metonymically (the sound for the movement).

47The extension of the denotative function of the sign thus created from the jaw(s) to the knee(s), to which homonymic *ĝenu- ‘jawbone, chin’ and *ĝenu- ‘knee’ may bear witness, would have entailed the loss of the self-referential nature attributed to the core invariant <ĝ> in the root noun *ĝenu, since the latter would no longer have referred back to its purported mandibular origins. This loss would have considerably enhanced the arbitrary nature of the word, even before it became fully conventionalized. Once the knee(s) had been named in this way (*ĝenu- ‘knee’), verbal roots denoting goal-orientated knee actions such as (of women) bending knees when begetting a child (cf. *ĝenh1- ‘to beget a child’) as well as the result of the action of begetting (cf. *ĝenh1- (also) ‘to be born’) could have been derived morphologically (cf. *-h1-) from the canonical form *ĝen- that I hypothesize was common to the root noun *ĝenu and the verbal root *ĝenh1- at some early stage of the language system that developed into PIE.

48As for the possible relation between *ĝenh1- ‘to beget a child, be born’ and *ĝneh3- ‘to know, recognize’ evoked by Meillet (1927) and others, it may be pointless to add to the speculation. And yet two or three aspects seem worthy of note: firstly, *-eh3- in *ĝneh3- may be a later, suffixal form of a canonical root *ĝen- (Southern, personal communication, 17.3.2005). Secondly, as Sweetser observes (1990: 18), “more abstract domains of meaning tend to derive their vocabulary from more concrete domains (rather than vice versa)”; in particular, physical-domain verbs frequently come to have mental-state meanings, e.g. ‘sight’ (physical vision) > ‘knowledge’ (mental ‘vision’). And thirdly, the English verb know does, in one of its biblical connotations, have a meaning related to the act of intercourse, which may lead to the birth of a child, as LeGates points out

“A second difficulty [in the Judeo-Christian tradition], was that women’s speech was connected to sexual desire because the serpent had spoken to Eve in the Garden of Eden in order to seduce her and she in turn had spoken to Adam. The result of all this wanton conversation was original sin. Thus knowledge had a sexual connotation: to know a woman meant to know her sexually”. (2001: 59)

49STEELS also provides a possible explanation for certain phenomena observable in the body-part vocabulary of non Indo-European languages that may also result, mutatis mutandis, from the kind of self-mimetic naming strategy posited herein. Whereas these phenomena appear to have become largely obscured in the Indo-European languages due to external factors such as taboo, and linguistic processes such as phonetic change, metathesis, reanalysis, etc., they remain more or less transparent elsewhere. For instance, in Pocomchi (Mayan), the term for ‘knee’ is nah ch’ehk (‘head of lower leg’), while in Hausa (Moghano), the term for ‘elbow’ is gwiwar hannu (‘knee of arm’), according to Andersen (1978: 355).

50Furthermore, most of the linguistic studies that I mentioned at the beginning of this paper date from the earlier part of the twentieth century, as if the subject were no longer worthy of serious attention. And yet scholars quoted above, such as M. Bloomfield, Buck, and Cuny, clearly expressed their intuition that the abstract notions which characterize parts of the human body, the bodies of animals, and manufactured instruments denoted by the same term in certain Indo-European languages may in some way have conditioned the term in question and the grammatical processes undergone by it.

51The epistemological problem posed by STEELS as a theory of the emergence of the sign is primarily one of boundaries: by conception and by definition, the linguistic sign is arbitrary, i.e. externally unmotivated. Its value within a given system is determined differentially on the syntagmatic and paradigmatic axes by its simultaneous relation to other, like signs in the system (position, function, distribution). By this view – and it is a view according to Saussure, since he remarks that in linguistics, “one might say that it is the viewpoint adopted which creates the object” (1983: 8) –, it is inconceivable that the linguistic sign should be motivated by system-external factors, except in ‘marginal’ cases such as onomatopœia, although it may of course be internally motivated to a variable degree. However, an analysis of submorphemic components endowed with a potential for meaning, such as phonæsthemes, suggests that while the linguistic sign is arbitrary, some of its sublinguistic components may have become arbitrary.

52Whatever the case, the semiogenetic approach adopted herein points to the conclusion that the hypothesized existence, at an indeterminably early stage in the evolution of Homo Sapiens, of an unconscious, somatotopically grounded naming strategy based on self-referentiality and conceptual transfer, provides a plausible explanation for the emergence, hundreds of millennia later, in the relatively recent form of language known as PIE, of an identical form denoting two body-parts that share a large number of physiological/kinematic properties and functions. Even though it may never be possible to reconstruct the original, prelinguistic form of this sign, we can at least accept the likelihood, thanks to the various techniques of reconstruction at our disposal, that in the Proto-Indo-European roots denoting the jawbone(s) or the chin on the one hand, and the knee(s) on the other, the linguistic form it had assumed was identical: *ĝenu-.

Haut de page


Andersen, E.S. 1978. Lexical universals of body-part terminology. In J. Greenberg (ed.), Universals of Human Language, vol. 3, Word Structure. Stanford: Stanford University Press, 335-368.

Bear, M.F., B.W. Connors & M.A. Paradiso 1996. Neuroscience: Exploring the Brain. Baltimore: Williams & Wilkins.

Benveniste, E. 1927. Un emploi du nom du ‘genou’ en vieil-irlandais et en sogdien. Bulletin de la Société de Linguistique de Paris 27: 51-53.

Benveniste, E. 1964 [1939]. Répartition des consonnes et phonologie du mot. In Études phonologiques dédiées à la mémoire de M. le Prince N.S. Trubetzkoy. Alabama: University of Alabama Press, 27-35.

Bergen, B.K. 2004. The psychological reality of phonæsthemes. Language 80-2: 290-311.

Bloomfield, M. 1891. On adaptation of suffixes in congeneric classes of substantives. American Journal of Philology XII: 1-29.

Bohas, G. 2000. Matrices et étymons. Développements de la théorie. Lausanne: éditions du Zèbre.

Browman, C.P. & L. Goldstein 1992. Articulatory phonology: an overview. Phonetica 49: 155-180.

Buck, C.D. 1949. A Dictionary of Selected Synonyms in the Principal Indo-European Languages. Chicago: University of Chicago Press.

Cahen, M. 1927. ‘Genou’, ‘adoption’ et ‘parenté’ en germanique. Bulletin de la Société de Linguistique de Paris 27: 56-67.

Chantraine, P. 1999 [1968]. Dictionnaire étymologique de la langue grecque. Paris: Klincksieck.

Corballis, M.C. 2003. From hand to mouth: the gestural origins of language. In M.H. Christiansen & S. Kirby (eds.), Language Evolution. Oxford: Oxford University Press, 201-218.

Cuny, A. 1924. Études prégrammaticales sur le domaine des langues indo-européennes et chamito-sémitiques. Paris: Librairie ancienne Honoré Champion.

Damasio, A. 2003. Looking for Spinoza. Joy, Sorrow and the Feeling Brain. London: Heinemann.

Dobson, E.J. 1968 [1957]. English Pronunciation 1500-1700, 2nd ed., 2 vols. Oxford: Clarendon.

Donald, M. 2001. A Mind So Rare. The Evolution of Human Consciousness. New York and London: W. W. Norton.

Ernout, A. & A. Meillet 2001 [1932]. Dictionnaire étymologique de la langue latine, reprint of the revised 4th ed. Paris: Klincksieck.

Fauconnier, G. & M. Turner 2002. The Way We Think. Conceptual Blending and the Mind’s Hidden Complexities. New York: Basic Books.

Feldman, J.A. & S. Narayanan 2004. Embodied meaning in a neural theory of language. Brain and Language 89: 385-392.

Firth, J.R. 1930. Speech. London: Ernest Benn.

Gallese, V. 2004. Intentional attunement. The Mirror Neuron system and its role in interpersonal relations. Interdisciplines (European Science Foundation),

Gallese, V. & G. Lakoff 2005. The brain’s concepts: the role of the sensory-motor system in conceptual knowledge. Cognitive Neuropsychology 22 (3/4): 455–479.

Gärdenfors, P. 2000. Conceptual Spaces. The Geometry of Thought. Cambridge (Mass.): MIT Press.

Gentilucci, M., F. Benuzzi, M. Gangitano & S. Grimaldi 2001. Grasp with hand and mouth: a kinematic study on healthy subjects. Journal of Neurophysiology 86: 1685-1699.

Heine, B. 1997. Cognitive Foundations of Grammar. New York: Oxford University Press.

Hilmarsson, J. 1982. Indo-European ‘tongue’. Journal of Indo-European Studies 10: 355-367.

Hodge, C.T. 1997. An Egyptian Etymology: Egypto-Coptic m}č. Anthropological Linguistics 39, n° 2 (Summer): 196-219.

Hurford, J.R. 2003. The Language Mosaic and its Evolution. In M.H. Christiansen & S. Kirby (eds.), Language Evolution. Oxford: Oxford University Press, 38-57.

Jóhanesson, A. 1949. Origin of Language: Four Essays. Reykjavík: H. F. Leiftur.

Johnson, M. 1987. The Body in the Mind. Chicago: University of Chicago Press.

Joly, A. & D. O’Kelly 1990. Grammaire systématique de l’anglais. Paris: Nathan.

Koch, W.A. 2001. Consciousness, communication, speech: a condensed view of the origins of language.

Koch, W.A. 2008. The Iconic Roots of Language. Lüdenscheid: RAM-Verlag.

Ladefoged, P. 2001. Vowels and Consonants, 2nd ed. Oxford: Blackwell.

Lakoff, G. 1987. Women, Fire, and Dangerous Things. What Categories Reveal about the Mind. Chicago: University of Chicago Press.

Lakoff, G. 1990. The invariance hypothesis: is abstract reasoning based on image-schemas? Cognitive Linguistics 1-1: 39-74.

Lakoff, G. & M. Johnson 2003 [1980]. Metaphors We Live By. Chicago and London: University of Chicago Press.

LeGates, M. 2001. In Their Time: A History of Feminism in Western Society. London: Routledge.

Lehmann, W.P. 1993. Theoretical Bases of Indo-European Linguistics. London: Routledge.

Loth, J. 1923. Le mot désignant le genou au sens de génération chez les celtes, les germains, les slaves, les assyriens. Revue celtique XL: 143-152.

MacNeilage, P.F. 1998. The frame/content theory of evolution of speech production. Behavioral and Brain Sciences 21: 499-546.

MacNeilage, P.F. 2008. The Origin of Speech. Oxford: Oxford University Press.

Mallory, J.P. & D.Q. Adams 1997. Encyclopedia of Indo-European Culture. London and Chicago: Fitzroy Dearborn.

Martinet, A. 1957. Phonologie et ‘laryngales’. Phonetica 1: 7-30.

Martinet, A. 1986. Des steppes aux océans. Paris: Payot.

Matisoff, J.A. 1978. Variational Semantics in Tibeto-Burman : The “Organic” Approach to Linguistic Comparison. Philadelphia: Institute for the Study of Human Issues.

Meillet, A. 1927. Lat. genuīnus. Bulletin de la Société de Linguistique de Paris 27: 54-55.

Meunier, F. 1871. La racine indo-européenne gan. Bulletin de la Société de Linguistique de Paris I: 73-74.

Morris, D. 2002. Peoplewatching [1997 under the title Manwatching]. London: Vintage.

Onians, R.B. 1988 [1951]. The Origins of European Thought about the Body, the Mind, the Soul, the World, Time, and Fate. Cambridge: Cambridge University Press.

Onions, C.T. (ed.) 1966. The Oxford Dictionary of English Etymology. Oxford: Clarendon Press.

Paget, (Sir) R. 1930. Human Speech. London: Kegan Paul, Trench, Trubner & Co.

Peirce, C.S., 1991 [1906]. Prolegomena to an apology for pragmaticism. In J. Hoopes (ed.), Peirce on Signs. Chapel Hill: The University of North Carolina Press, 249-252.

Philps, D. 2006. From mouth to hand. In A. Cangelosi, A.D.M. Smith & K. Smith (eds.), The Evolution of Language. Singapore: World Scientific Publishing, 247-254. On-line:

Pokorny, J. 1959. Indogermanisches etymologisches Wörterbuch, 2 vols. Bern and Munich: Franke.

Rix, H. (ed.), M. Kümmel, T. Zehnder, R. Lipp & B. Schirmer 2001. Lexikon der indogermanischen Verben. Wiesbaden: Reichert.

Rizzolatti, G. & L. Craighero 2004. The mirror-neuron system. Annual Review of Neuroscience 27: 169-192.

Rizzolatti, G. & C. Sinigaglia 2008. Mirrors in the Brain. How Our Minds Share Actions and Emotions, translated by Frances Anderson. Oxford: Oxford University Press.

Saussure, F. de 1983. Course in General Linguistics, translated and annotated by Roy Harris. London: Duckworth.

Searle, J.R. 1983. Intentionality. An Essay in the Philosophy of Mind. Cambridge: Cambridge University Press.

Southern, M.R.V. 1999. Sub-Grammatical Survival: Indo-European s- mobile and its Regeneration in Germanic. Washington DC: JIES 34 / Institute for the Study of Man.

Sweetser, E.E. 1990. From Etymology to Pragmatics. Metaphorical and Cultural Aspects of Semantic Structure. Cambridge: Cambridge University Press.

Szemerényi, O.J.L. 1999. Introduction to Indo-European Linguistics. Oxford: Oxford University Press.

Trask, R.L. 1996. A Dictionary of Phonetics and Phonology. London: Routledge.

Wallis, J. 1653. Grammatica Linguæ Anglicanæ, Oxford, reprint n° 142 (1969), R.C. Alston (ed.). Menston (Yorkshire): The Scolar Press.

Watkins, C. 2000. The American Heritage Dictionary of Indo-European Roots, 2nd ed. Boston: Houghton Mifflin.

Wilkins, D.P. 1996. Natural tendencies of semantic change and the search for cognates. In M. Durie & M. Ross (eds.), The Comparative Method Reviewed: Regularity and Irregularity in Language Change. New York: Oxford University Press, 264-304.

Haut de page


1 See Martinet 1957.

2 See Bear, Connors, & Paradiso 1996: 413-415.

3 See Morris 2002: 364-370 for the notion of body self-mimicry.

Haut de page

Pour citer cet article

Référence électronique

Dennis Philps, « Conceptual transfer and the emergence of the sign »CogniTextes [En ligne], Volume 2 | 2009, mis en ligne le 10 décembre 2009, consulté le 01 décembre 2023. URL : ; DOI :

Haut de page


Dennis Philps

Université de Toulouse-Le Mirail (IRPALL-CAS)

Haut de page

Droits d’auteur


Le texte seul est utilisable sous licence CC BY-NC-ND 4.0. Les autres éléments (illustrations, fichiers annexes importés) sont « Tous droits réservés », sauf mention contraire.

Haut de page
Rechercher dans OpenEdition Search

Vous allez être redirigé vers OpenEdition Search