1This second issue of Lexis is devoted to the topic of lexical submorphemics, an emerging – and therefore still controversial – field of study whose aim is to account for elements in the lexicon of a language that form parts of morphemes, and are construed as playing a role in the construction of ‘meaning’. Its object of study is the ‘submorpheme’: its essence, structure, function, combinatorial properties, distribution, invariance and variability, as well as its links with notions, cognition, motricity, Cratylism, language acquisition, etc., whether synchronic, diachronic, or panchronic in nature. Arguably, one of the strong points of this field in terms of other, more traditional approaches to the sign, is its ability to identify and systematise possible links between the linguistic and the extralinguistic, a feature that lends pertinence to its contribution to the scientific study of language as a whole.
2However, the history of the development of modern linguistics has been such that even today, defining the concept of ‘submorpheme’ satisfactorily is no easy task. If one takes the definition found in Crystal’s Dictionary of Linguistics & Phonetics, i.e. “a part of a morpheme that has recurrent form and meaning, such as the sl- beginning of slimy, slug, etc.” (Crystal 2003: 301), for example, one may be forgiven for concluding that the nature of the submorpheme is exclusively lexical. And yet, as will be seen, this concept is also required in grammar, since a number of English function words may be also be segmented into smaller units by adapting a principle of invariance, first set out by the French linguist Guillaume (1883-1960), to the submorphemic level. Guillaume expresses this principle in the following way: “[T]he linguistic sign establishes an invariant condition in tongue [Fr. langue, see Hirtle’s introduction to Guillaume 1984] from which there develops an infinite number of consequences” (1964: 150, my translation).
3Although Guillaume never did so himself, transposing this principle from the morphemic to the submorphemic level brings to light the fact that word-initial submorphemes displaying both phonological and notional invariance, such as the sl- segment used as an example by Crystal in slack, slime, slop, slouch, slow and slug (see Tournier 1985: 146), are endowed not with meaning, but with a potential for meaning which requires certain conditions to be met, whether at system or discourse level, to become linguistically meaningful (see Philps 2005). The logic that underlies this transposition runs as follows: any value attributable to a given set of words on the basis of notional invariance (e.g. softness or slowness in English ‘sl- words’) must ultimately be located in the smallest structurally invariant segment of these words (e.g. sl- in slack, slime, slop, etc.), on the grounds that within these words, an invariant notion cannot, by definition, be conveyed by a variable segment (e.g. -ack in slack, -ime in slime, -op in slop, etc.).
4As we come towards the end of the first decade of the 21st century, establishing exactly what we mean by ‘submorphemics’ is decidedly no easy task. To do so requires, at the very least, a sense of history, and involves making a distinction between lexical submorphemics and grammatical submorphemics on the one hand, and, on the other, the complementary approaches to submorphemics adopted by linguists working within what I will term, albeit simplistically, the ‘Anglo-Saxon’ tradition and the ‘French’ tradition respectively. However, any distinction made initially between lexical submorphemics and grammatical submorphemics cannot be considered to be absolute, as a number of those function words whose etymology has been reconstructed, both in English and in other languages, are known to have a lexical origin, e.g. downadv. < late OEng. dúne, dún, aphetic form of adúne ‘adown’, weakened from OEng. of dúne ‘off the hill or height’ (Oxford English Dictionary, 2nd ed., 1989).
5In morphology, we may begin by recalling the comments made by Harris (1909-1992), a key mentor to Chomsky, on /D/ in there, then, this, etc., and /hw/ in where, when, which, etc.:
Difficulties of stating meaning also occur in the case of morphemes of unique environment, which are segmented off when an utterance has a unique residue after all morphemes have been divided off. […] in there, then, thither, this, that, etc., we obtain […] a segment /D/ with demonstrative meaning, plus various residue elements with unique meanings (/is/ ‘near’, /æt/ ‘yonder’, etc.). In where, when, whither, which, what, why, etc., we obtain an element /hw/ with interrogative (and in some linguistic environments, relative or resumptive) meaning, plus various second elements with unique meanings (/at/ ‘object reference’, /ay/ ‘reason’, etc.). (1951: 192-193)
6In the pre-war and post-war periods, the term ‘submorphemic’ did not have the meaning it has today, as is clear from the following remark made by Hockett:
The statement as to the occurrence of the morpheme i―to which no meaning can be assigned―will have to operate in terms of submorphemic (phonemic) properties of environments, whereas on the tactical level we should like to be able to state environments of occurrence and non-occurrence of classes of morphemes in terms of other classes of morphemes, without regard to submorphemic matters.” (1947: 332)
7Forty years after Harris’s comments, the French linguists Lapaire & Rotgé, working within an ‘enunciative’ framework (see also Danon-Boileau 1983 and Delmas 1987), assert that within the class of grammatical ‘th- words’ in English,
THIS and THAT both contain the morpheme TH-, which reflects a set of mental operations preceding the speech act that is both diverse and complex, and on whose existence the speaker counts […]. Hence TH- endows THIS/THAT with its ‘memorial’ dimension, or, put differently, its ability to target the past.” (1991: 62, my translation).
8Even today, one can still read that the and that “both contain the morpheme TH-, conveying the notions of ‘operational precedence’ (Lapaire & Rotgé), of ‘phase 2/thematicity’ (Adamczewski), of ‘pinpointing’ (Culioli), in short, of ‘alreadiness’.” (Jamet et al., 2007: 75, my translation). But is TH- a morpheme? The same question may be asked of WH- in grammatical ‘wh- words’. If it is accepted that morphemes combine with other morphemes to form words, then TH- and WH- in ‘th-’ and ‘wh-’ words respectively must be analysed as submorphemes, since they combine not with morphemes, but with submorphemes (e.g. -ere in there, -en in when, etc.), although not all grammatical items can necessarily be segmented in this way.
9To return to lexical submorphemics, it must be said that the subject has always provoked controversy, as Bottineau (this issue) points out:
The question of lexical submorphemics has long been controversial in English lexical studies and in general […]. The linguistic sign is traditionally considered as ‘arbitrary’: the phonological form is not iconic and does not reflect the physical properties of the object or event to which the word refers […]. On the other hand, a minority of authors are convinced that some areas of the English lexicon are organized by a paradigm of submorphemic markers: clusters of more than one phoneme, and occasionally isolated phonemes, which appear to relate the lexical notion to a specific semantic field. Between these two positions there is hardly any room for scientific debate.
10The same can be said of those segments known as ‘phonæsthemes’ since Firth first coined the term in 1930. Phonæsthemes are frequently recurring, systematic pairings of form and meaning whose morphemic status is not clearly established (Bolinger 1950: 30, Bergen 2004: 290, etc.). In English, some of these pairings were noted as far back as the 17th century by Wallis (1653), in the first edition of his Grammatica Linguæ Anglicanæ, e.g. (word-initially) : st- / fixedness (in steady, stout, sturdy, etc.), wr- / torsion (in wrench, wring, wrist, etc.), and str-/ power (in strength, strive, struggle, etc.). Although phonæsthemes may well develop symbolically in an arbitrary manner (cf. gl- in the ‘gl- words’ related to vision and/or light below, for instance), this does not preclude the hypothesis that their initial motivation was partly iconic. As to their cognitive status, Bergen points out that this remains controversial “in part because they [phonæsthemes] do not fit well into linguistic theories that view compositionality as a defining characteristic of morphological complexity” (2004: 291).
11As bar-Lev reminds us (2005-2006 : 25), the American linguist Bolinger appears to have been the first scholar in the course of the 20th century to have theorised phonæsthemes and other submorphemes in the lexicon:
MORPHEMES AND SUB-MORPHEME DIFFERENTIALS. […] Where, within or between morphemes and mere differentials, would our rimes and assonances fall? If we can show enough regularity in use, a rime or an assonance should be, or come very near to being, a morpheme. Let us take the form /gl/ already noted as referring to ‘visual phenomena’, and assay its possibilities as a morpheme. Discarding technical, learnèd, and dialectal words, we may list, in seven columns showing graduating fidelity to the meaning ‘visual phenomena’, all the base words, excluding obvious derivatives, that begin with /gl/: [glance, glare, gleam, glimmer, glimpse, etc.]. (1950: 131).
12Among other topics, lexical submorphemics addresses areas such as sound symbolism (see Jespersen 1922, Fónagy 1983, Malkiel 1990, Hinton et al. 1994, Genette 1995, Humboldt 1999, and Chadelat, this issue), submorphemic iconicity (see Mannheim 1991, Waugh 1993, Waugh & Newfield 1995, Nuckolls 1996, and Philps, this issue), word-formation (see Adams 1973, Bauer 1983, Plag 2003, and Fandrych, this issue), and cognition (see Bottineau 2002, Philps 2006, as well as Argoud and Bottineau, this issue).
13As surely as this attempt to establish what is meant by ‘submorphemics’ requires a sense of history, so does it require a sense of the future. And indeed, whether lexical or grammatical, one particular aspect of submorphemics whose future seems assured is the gateway it provides into the field of cognition, as is foreshadowed by the term ‘conjecture’ in the title of Bottineau’s cognitively orientated paper, at least in the oldest, now obsolete sense of the word, namely “a conclusion as to coming events drawn from signs…” (OED). In this perspective, which, need it be mentioned, is only one of many, submorphemics doubtless holds the key to redefining the relation between speech and the embodied nature of language, both in time and in space, as well as that between speech and ideation. As such, the field will likely continue to breed controversy, a small price to pay, however, for the breakthroughs it promises in terms of deepening our understanding of the workings of the mind as observed through language.
14To turn to the papers published here, Didier Bottineau (CNRS, MoDyCo – University of Paris 10 Nanterre) focuses on the vowel and consonant clusters known as submorphemes, ideophones and phonaesthemes found in the English lexicon. The author investigates the cognitive relevance of submorphology for language users and observers, and attempts to determine whether such relevance exists at all, or whether the linguist is the only one to perceive a link between submorphemic order and lexical-semantic classes. Various aspects of the question are explored, including the nature of submorphemic semantics and its differentiated distribution.
15Line Argoud (PhD student, Grenoble - Université de Toulouse-Le Mirail) explores English lexical items beginning in bl- with a two-fold objective. Firstly, she presents the hypothesis that the phonological invariance of this heuristically-formed class corresponds to a notional invariance within two important subsets of ‘bl- words’, and that this notional invariance may be traced back to mental representations whose underlying source may be accounted for by Lakoff & Johnson’s <CONTAINER> schema. Secondly, she assesses the part played by aktionsart in the data provided by the entries for ‘bl- words’ culled from two dictionaries, the Oxford English Dictionary and Wright’s English Dialect Dictionary.
16Focusing on an inventory drawn from a current etymological dictionary, Jean-Marc Chadelat (IUFM de Paris) analyses the class of words defined by the word-initial sublexical marker <Cr-> in order to account for the motivation which is perceived by users of the language on the phonological level. His underlying hypothesis is that a semiogenetic continuum exists between the onomatopoeic and phonæsthemic dimensions of this marker by virtue of synesthetic transfer, akin to a process of conceptual transfer that appears to operate through the unconscious projection of typical features from one extralinguistic domain to another by means of a symbolic transposition which is at variance with the axiomatic arbitrariness of the linguistic sign.
17Ingrid Fandrych (National University of Lesotho) analyses the role played by elements below the morpheme level in word-formation processes, and reminds us that mainstream word-formation is concerned with the creation of new words from morphemes. Insofar as morphemes are fully linguistic signs, the resulting neologisms are transparent: speakers can deduce the meanings of the new formations from the meanings of their constituents. Thus, morphematic word-formation processes can be analysed in terms of their modifier/head relationship: A + B > AB. There are, however, certain word-formation processes that are not morpheme-based, and that do not have a modifier/head structure. The author claims that to account for words formed in this way, one is required to posit the existence of elements below the morpheme level.
18Lastly, Dennis Philps (University of Toulouse-Le Mirail) submits that the word-initial segment gn- (phonologically /n/) in the heuristically set up class of English ‘gn- words’ functions as a ‘sublexical marker’ (<gn->). He claims that although <gn-> must be considered as structurally indivisible in Modern English, <g> in <gn-> may originally have functioned as a ‘core invariant’ and <n> as a variable in the pre-history of English, notably in Indo-European. Analysing gn- in one major subset of ‘gn- words’ as the surface trace of an underlying notional domain linked to jaw-related phenomena (sound articulation, mastication, manducation, etc.) leads him to suggest that there exists an extralinguistically motivated, causal relation between <gn-> and this domain involving the concept of ‘submorphemic iconicity’.