Semic verbalisation: A systematic procedure for the replica- tion of meaning in translating lexical items

The purpose of this article is to introduce a systematic translation procedure called Semic Verbalisation (SV), which draws on componential analysis so as to first lexicalise and then translate a word's sense components. While construing a hybrid denotational-mentalistic conception of meaning, a series of steps will be formulated that are useful for solving various prac-tical translation problems. The procedure is based on philosophy of lan-guage (Russell, Tarski, Frege, Quine and Wittgenstein), linguistics (Katz & Fodor) and translation studies (in particular Newmark). Its practical appli-cations for translation studies will be illustrated, and, in the discussion, its advanta ges and drawbacks will be scrutinised from both a theoretical and a practical perspective


Introduction
Semic Verbalisation (SV) is a systematic procedure favouring the replication 1 of semantic features in translating lexical items.Through three related operations applied on a word's sense components (viz.determination, lexicalisation and translation), SV is intended to cater for semantic correspondence in translation praxis.Although SV is based on the Componential Analysis (CA) framework pioneered by Katz and Fodor (1964), 2 its conceptual foundations combine notions stemming from diverse semantic theories.Denotational and mentalistic conceptions of meaning will be assessed and integrated so as to formulate a workable semantic thesis allowing specific translation problems to be dealt with satisfactorily.As will become evident in the following pages, CA is presently taken simply as a procedure or analytical tool to determine what the meaning components of an expression may be, but the nature and origin of those components will be accounted for in terms of a hybrid denotationalmentalistic conception of meaning rooted in the philosophy of language.
More generally, SV seeks to offer practising translators the chance to substantiate practical decisions in non-intuitive terms and thus authenticate their choices in translating lexical items.For this purpose, Newmark's conception of the interface between Componential Analysis and translation (1988) will be scrutinised and taken as a basis to rationally account for the process of SV.
The paper is organised as follows: Section 2 will present the conception of meaning adopted for SV.In Sections 3, 4 and 5, the notion of CA will be reviewed in relation to its role in translation.Section 6 will introduce the process of SV and show its usefulness in solving specific translation problems.Finally, in Section 7, the advantages and disadvantages of SV will be discussed and general conclusions will be drawn regarding its practical and theoretical status.

The hybrid semantic foundations of Semic Verbalisation
Despite centuries of investigation and speculation, meaning remains a most elusive concept.No sooner did a theory provide an apparently satisfactory definition of meaning than another one emerged to prove it mistaken or incomplete. 3So far, not a single theory has escaped this ominous fate.
In contemporary terms, the first widely spread theory of meaning was the Denotational Theory of Meaning (cf.Russell, 1905;Tarski, 1944), which posited that the meaning of an expression is the object it denotes in the non-linguistic world.The explanatory deficiencies of this theory soon became evident (see Frege, 1952Frege, and, especially, 1892)).On the one hand, such a definition of meaning implies that if an expression is meaningful, it must necessarily have a physical denotation.This claim is obviously false, as it would entail that expressions of the likes of Pegasus, the, nothing and hello are devoid of meaning.On the other hand, the above definition grants that two expressions which have the same referent must be synonymous.This claim is also false, since, as shown in a now famous example by Frege (1892), the expressions the morning star, the evening star and Venus all denote the same object but they are not synonyms.Ultimately, meaning cannot be equated with denotation in these terms.
In response to the inconsistencies of the Denotational Theory, there emerged so-called Mentalist Theories of Meaning (e.g. de Saussure, 1916de Saussure, /1959;;Glucksberg & Danks, 1975), which maintain that "the meaning of each expression is the idea (or ideas) associated with that expression in the minds of speakers" (Akmajian et al. 2001: 233, see also Katz (1964) for a concise account of the rationale behind mentalism in linguistics).The main problem facing this conception of meaning is that the very notion of 'idea' is either so vague that the theory cannot explain or predict anythinghence becoming not testable-or so precise that the theory leads to false predictions (Akmajian et al, 2001: 233).Correlating meaning to mental images might help us explain the meaning of Pegasus, but when thinking of words like car people visualise the image of a specific species or type (for example, a sports car or a Mercedes), thus excluding the totality from their conception.This theoretical roadblock could be avoided by proposing that meanings are not actually ideas, but concepts, or mentally represented categories of things.However, as noted by Quine, meanings are evidently intended to be ideas, somehow -mental ideas for some semanticists, Platonic ideas for others.Objects of either sort are so elusive, not to say debatable, that there seems little hope of erecting a fruitful science about them.(1951: 22) Strictly speaking, therefore, ideas cannot be meaning, for the former are subjective, whereas the latter is objective (Frege, 1892).Several other semantic theories were created in an attempt to circumvent this problemfor example, the Sense Theory of Meaning (Frege, 1892) and the Use Theory of Meaning (Wittgenstein, 1922(Wittgenstein, ,1953))-but their elucidation escapes the scope of this paper.At this point, it should suffice to say that no semantic theory has yet been able to unequivocally define the notion of meaning (cf.Akmajian et al, 2001: 236).
For the sake of practicality, this paper will put forth a conception of meaning which combines denotation and mental representation.The term denotation will be used not only to signify the faculty of words to refer to actual objects, but also to denote referentiality to non-physical verbal and pragmatic facts or events.In this sense, the sustainment of referentiality relies on the consensual grammatical bases which allow for the communication of non-physical realities.For instance, the denotation of the word hello will be understood as the coincidence among language users on the exegetic apprehension (in verbal and/or pragmatic terms) of any particular instance of use of that word.
In addition, this paper will assume that, in spite of the plausible vagueness and/or excessive specificity of mental imagery, the similarities holding among the innumerably different mental representations of a word are enough to ensure that its meaning, in a broad sense, will be similar for the members of a speech community.In brief, this paper posits that a denotational-mentalistic definition of meaning in the above terms is solid enough to support the construction of a rational tool aimed at solving specific translation problems.

Componential Analysis, definition and synonymy
For the purposes of this paper, Componential Analysis or CA (Katz & Fodor 1964) will be defined as a mechanism through which words can be dissected so as to reveal their constituting sense components.Within this framework, the different single senses of a word are termed sememes, and their semantic features are given the name of semes.For instance, the word salt might be said to comprise two different sememes, the semes involved in each of those sememes differing radically.For example, salt in a paper on chemistry will comprise semantic features such as [+SOLID], [+CRYSTALLINE] and [+MINERAL], whereas when used to describe a member of a ship's crew, salt will include the semes [+SAILOR], [+OLD] and [+EXPERIENCED], to name but a few.Consider the following decontextualised example to see what a basic instance of CA looks like: In the light of this analysis, the meaning of the sememe cry could be said to be loud, vocal sound.Thus, the expressions cry and loud, vocal sound could be deemed synonymous.Additionally, the expression loud, vocal sound could be seen as the definition of the expression cry.
Definitions of any kind presuppose the existence of some sort of previous knowledge.Language users do not need to be well versed in semantics to tacitly acknowledge the conventional nature of the linguistic sign and employ it effectively in communicative acts, just as they do not need to know what a word precisely means to take advantage of it as a device through which to fulfil their communicative goals.Whenever that goal implies providing a definition, language users will be unconsciously relying on CA, as may be gathered from the following lines by Quine: Definition is not, indeed, an activity exclusively of philologists.Philosophers and scientists frequently have occasions to "define" a recondite term by paraphrasing it into terms of a more familiar vocabulary.But ordinarily such a definition, like the philologist's, is pure lexicography, affirming a relationship of synonymy antecedent to the exposition in hand (1951: 24-5).
By establishing a correlation between definition and CA, it could be claimed that, to the extent that language users (in our case, translators) are communicatively competent, they do not need to be philologists or specialists to engage in successful instances of CA.All language users may resort to all words, including technical ones, to fulfil their communicative goals, even if they do not fully grasp the actual 'meaning' of the word, that is, even if they are not aware of all the semes the word involves.
Let us now turn to synonymy and discuss its relation to CA, a task in which philosophical difficulties are anything but milder.There is no solid elucidation that accounts for what the grounds for synonymy are, and the prospects of a definite explanation are not too good.Yet, this does not imply that there is no hint that may pave the way for fruitful findings: Just what it means to affirm synonymy, just what the interconnections may be which are necessary and sufficient in order that two linguistic forms be properly describable as synonymous, is far from clear; but, whatever these interconnections may be, ordinarily they are grounded in usage.Definitions reporting selected instances of synonymy come then as reports upon usage (Quine 1951: 25).
Put succinctly, synonymy, as well as CA, might be said to be rooted in the apprehension of actual language use.Even in the absence of a theory that unmistakably explains what the interconnections supporting synonymy may be, and of a theory that indisputably explains what semes are or how they can be determined, both synonymy and CA are put at work in daily language use, and their very manifestation might be enough for language users to share the stakes which make them useful, practical and communicatively effective.
The above argumentation has two noteworthy implications for translation praxis: firstly, it implies that translators may overlook deep philosophical speculation, which does little to explain the nature of definitions, synonymy or CA.Secondly, it entails that translators' mere contact with language is enough to help them develop an intuition that will enable them to efficiently use definitions, synonymy and, in our case, CA.
Yet a problem of a professional and/or academic nature prevails, since translators, as members of a professional and scholarly community, usually do not invoke intuition as the sole source of their operative decisions.The following sections deal with how CA can be used to sort out practical translation problems and to validate choices that otherwise could be explained only in intuitive terms.The foundations for SV will thus be laid.

Towards semic verbalisation: building upon Newmark
Not many translation scholars have shown themselves as positive about the benefits of CA in translation as Peter Newmark.His confidence on its efficiency is such that he has described it as "the most accurate translation procedure, which excludes the culture and highlights the message" (Newmark, 1988: 96).Although such a claim may be a little too extreme, Newmark is far from wrong in hinting at the usefulness of CA -other translation scholars, such as Nida (1975), Holzhausen (1981) and Vossoughi (1996), have also discussed the applications of CA.In addressing its use in translation, Newmark explains that the basic process is to compare an SL [Source Language] word with a TL [Target Language] word which has a similar meaning, but is not an obvious one-to-one equivalent, by demonstrating first their common and then differing sense components.(1988: 114) For Newmark, the application of CA in translation requires that an SL sememe and a TL sememe presumed to be equivalent be analysed individually in their respective languages, so that their degree of equivalence can be assessed by comparing their constituting semes.Such a procedure, it must be noted, seems to have been formulated with an ideal translator in mind.It requires translators to know, in advance, the full array of semes included in an SL item and in its most ready TL equivalent.In this sense, Newmark's conception of the use of CA in translation does not seem to provide any valuable insights into how the semantic dissection of a lexical item might aid translation practice.However, it does represent a good starting point on the basis of which to develop some practical strategies for sorting out particular translation problems.
On the assumption that a translator's semantic intuition of an SL lexical item is enough to ensure that the translator will correctly assess its communicative value and function within a specific context, 4 CA could be turned into an advantage, not on the comparative premises proposed by Newmark, but as an analytical device through which the ST (Source Text) semes to be replicated in the TT (Target Text) can be determined, after which the translator will choose the most preferable way of replicating such semes.
It must be noted, at this point, that semes are not word-particular.The words shriek and shout, for instance, share the seme [+LOUD], among many others.Consequently, the semantic load of any given lexical item can be realised by using a hyperonym and combining it with the metalanguage of its own semes.Consider these simplified, decontextualised examples: On the basis of such componential analyses, it could be claimed that, adhering to the conception of meaning proposed in this paper, the sentence (1) John heard a shriek coming from the basement.has the same meaning as (2) John heard a high-pitched cry of anger coming from the basement.
The above example illustrates how CA can be used as an intralingual mechanism for achieving semantic correspondence between different expressions.Furthermore, it shows how semantic dissection could be turned into a helpful translation tool: an examination of the semes constituting the semantic configuration of an SL expression should cater for the replication of the same overall meaning in the TL, regardless of whether a one-to-one equivalent in the TL is available or desirable.Thus, if the following Spanish sememes were componentially analysed as follows: a sentence like (1) could be translated as either (3) Michael oyó un alarido proveniente del sótano.
[Literally: John heard a high-pitched cry of anger coming from the basement.]

A semic taxonomy
Although semes have been described and classified by several linguists and semioticians, no theory has yet been able to explain how they can be determined, or how many of them there are in any given sememe.
Objectively, no complete enumeration can be made of the semes contained in a sememe.Nevertheless, for practical purposes, a translator does not consider all of the semes present in a word.Yet, a crucial question remains: which are the semes a translator needs to be aware of?
Any attempt to answer this question should take into account the fact that semes are rarely considered in isolation.Just as words may be grouped into lexical fields (see Trier, 1931;Martin Migorance, 1984;Faber & Mairal, 1994), semes can be catalogued according to semantic classes.Within each class, two main types of semes can be differentiated: those indicating that a sememe belongs to a given semantic class are termed generic semes, whereas those distinguishing a sememe from others of the same class are referred to as specific semes.The former are classified into three kinds, namely mesogeneric, macrogeneric and microgeneric semes, which, in turn, correspond to particular semantic classes, respectively called domains (social contexts or fields of human activity), dimensions (general paradigms grouped into scalar, ternary or binary oppositions, such as [+CONCRETE] vs. [+ABSTRACT]), and taxemes (the minimal paradigms in which sememes are interdefined). 5 For example, the taxeme //UNDERWEAR// comprises four sememes.Each of them contains the microgeneric seme [+UNDERWEAR] and is distinguished from the other sememes of the same taxeme by a specific seme: [+USED OVER FEET] in socks, [+USED OVER GENITALIA] in underpants, [+USED OVER BREASTS] in bra, and [+USED OVER TRUNK] in undershirt.As this taxeme falls under the domain //CLOTHING//, the four sememes also contain the mesogeneric seme [+CLOTHING].Finally, the four sememes are also members of the common dimensions that define macrogeneric semes, such as [+INANIMATE]. 6Figure 1 below lists some of the main semes included in the sememe socks: Notice that not every individual would have similar conceptions of how to classify semes (for example, not everybody would agree in saying that socks are underwear).However, as has been explained in Sections 3 and 4, and as will be reemphasised below, the usefulness of CA as an analytical device lies not in its indisputability as an objective semantic classification tool, but in the systematicity that it offers translators seeking to substantiate their decisions in non-intuitive terms.
In translation, CA can be used to determine which SL semes are usually rendered into the TL.By establishing which taxeme an SL lexical item belongs to, and by contrasting it with the other sememes included in the taxeme, translators will be able to identify both the generic and specific semes of the lexical item in question.However, it is not necessary to replicate all of the semes identified in an SL expression.Generic semes may often be left untranslated, as they are usually implied by the context in which the lexical item occurs, or by the collocations which accompany the lexical item in a text.In fact, a microgeneric seme will always imply its corresponding mesogeneric seme, for example, the seme [+UNDERWEAR] will always imply the seme [+CLOTHING].Specific semes, on the other hand, carry the distinguishing semantic traits of a word, thus typically requiring more scrupulous replication -in the above example, what sets socks apart from other sememes included in the taxeme [+UNDERWEAR] is the specific seme [+USED OVER FEET]. 7

Semic Verbalisation
Once the semes requiring interlingual rendering have been determined, the translator may choose to replicate the identified semantic traits of the SL item by translating the 'label' or 'name' of each seme.This is what I have

SEMEME: Socks
termed SV, which could be formally defined as the process whereby an SL word is rendered into the TL by first determining, then lexicalising, and finally translating its constituting semes.
Lexicalisation, in this context, must be understood as the realisation of a semantic unit (a seme) as a lexical unit (a word); the underlying idea is that, even if the graphemic conventions for representing semes and words are almost identical, semes belong to a higher level of abstraction than words.Thus, the word underwear, for example, could be seen as the lexical label resulting from the lexicalisation of the seme [+UNDERWEAR], a more abstract unit.
In brief, a three-stage procedure allows translators to engage in SV: firstly, the SL word's semes will be determined by engaging in CA; secondly, those semes will be lexicalised, that is, taken from the realm of semantic representation to that of the lexicogrammar; 8 finally, the resulting lexical items will be translated into the TL.
The following are some of the cases in which SV might prove desirable, or even necessary, to sort out practical translation problems.

SV in translating cultural words
Since cultural words denote objects or concepts emerging within a specific culture, they simply do not allow for a 'direct' translation into another language.Take the Spanish word mate¸ for instance.As there is no English word to refer to this beverage, it is necessary to lexicalise the semes it contains for a target text reader to understand what that word means.Thus, if the word mate were componentially analysed as follows: An understanding of the processes of CA and SV is invaluable to these explanatory purposes, since CA and SV synthesise the formal underlying procedure operative in the clarification of cultural words.

SV in the absence of one-to-one equivalents
Words forming a lexical set in an SL seldom have TL 'equivalents' containing the same semes.This scenario is most evident when contrasting the different words making up hyponymic sets across languages.For example, if the word sneer were componentially analysed as comprising the following semes: 9

SEMEME: sneer
[Literally: The captain smiled scornfully at the rookie.] Once the translator has covered all the steps leading to a translation via SV, that is, the identification of the semes in an SL word, the lexicalisation of those semes in SL, the rendering of the resulting lexical items into the TL, the translator will be in a position to alter the resulting basic syntactic structure.For example, if the translator were not satisfied with the naturalness of sentence ( 9), the syntactic roles of the lexicalised semes could be reconfigured and other pertinent words could be included so as to produce sentences such as (10) El capitán sonrió al novato con desdén.
[Literally: The captain gave the rookie a scornful smile.] The possibilities of lexical and syntactic reconfiguration are actually endless, or at least as numerous as the translator's imagination or skills will allow for.Moreover, it should be noted that both lexical and syntactic reconfiguration as operations subsidiary to SV can be enacted in all cases where SV was previously used.

SV and the avoidance of cacophony
The use of a co-textually cacophonic TL one-to-one equivalent can be avoided by lexicalising a microgeneric seme and then narrowing down its meaning by lexicalising the specific and macrogeneric semes.Thus, if we assumed the words behold and contemplar to be one-to-one equivalents, and, recognising their specific usage in learned religious circles, analysed the TL word as comprising the following semes: By avoiding the co-occurrence of the words contemplen and templo, translation ( 14) succeeds in eliminating the repetition of the phonological sequence /templ/ in (13), which is cacophonic not just because of mere iteration, but also because the sequence itself contains two plosive sounds, namely /t/ and /p/, which logically multiply to four.

SV in the enactment of an interlingual tenor shift
When the translation task requires the translator not only to render the ST into the TL, but also to adjust the TT to a less specialised audience than that

SEMEME: contemplar
of the ST, it is usually necessary to 'explain' the meaning of field-specific words using everyday vocabulary in the TL.For instance, such is the case when an amateur who is keen on medicine asks for a medical paper to be translated, or when we are asked to translate a technical article from a journal so that it can be used in a secondary-school textbook.These translation acts require a change of tenor, that is, an adjustment of the relationship between the participants in a communicative act (Halliday 1985), in order to support the process of interlingual conversion.Thus, the translator is compelled to adjust any lexical choices to be made to the register of the target audience.
As the very process of SV consists of paraphrasing an expression by using more, less specific words, it seems adequate to enact the interlingual lexical modifications mentioned above.For instance, if we analysed the word pheochromocytoma as comprising the following semes: [Literally: The patient was diagnosed with pheochromocytoma.]could be translated for a TL secondary-school audience as (16) The patient was diagnosed with a small, benign adrenalinesecreting vascular tumour (derived from chromaffin cells).

SEMEME: hpeocromocytoma
It is worth noting that SV does not imply that all of the macrogeneric semes identified in a lexical item are an essential part of reduplication.In fact, the reproduction of macrogeneric semes will be necessary only when these prove crucial to validating the distinctiveness of specific semes.For what distinguishes pheochromocytoma from other tumours in the example above is not that it is the only tumour which is both 'adrenaline-secreting and derived from chromaffin cells' (specific seme), but that it is the only 'small, benign vascular' (macrogeneric semes) tumour to exhibit such specific characteristics.In other words, it is not microgeneric semes per se that are designated by specific semes, but rather microgeneric semes as described by macrogeneric semes.Moreover, the characteristics of the target audience may even allow for the omission of specific semes.In the example above, the specific seme [+DERIVED FROM CHROMAFFIN CELLS] may be disregarded as it is too intrinsically technical for a secondary-school audience to grasp or even care about, and scrupulousness in SV may compromise the consistency or global coherence of the target text.

Discussion and conclusion
At this point, two pertinent observations may have suggested themselves to the reader.The first observation is that, very frequently, translators and translation trainees successfully apply strategies similar to those involved in SV without even knowing that a process exists that may be called SV.The second observation is that the same target texts that may result from the application of SV could be produced intuitively, through 'common sense'.What, then, the reader may ask, is the usefulness of SV in the light of these two facts?
The answer is one and the same for both points.What SV does is to provide translators with a theoretically justifiable step-by-step procedure to translate lexical items.Hence, it should enable translators and translation trainees to develop a more solid scholarly profile -as it represents an academically feasible means to account for specific translation choicesand, thus, to professionalise their praxis since intuition cannot be the source of a professional translator's operative decisions.
However, SV evidences certain drawbacks.Above all things, it tends to result in wordiness.To overcome this problem, the translator may consider applying other complementary strategies, such as paraphrasing, précising or generalising either the target sentence or those surrounding it.Another drawback of SV is that translators can never be absolutely sure that their analyses of a sememe's semantic constitution are correct or even exhaustive.On this score, translators are advised to be judicious in deciding when and where to use SV, and to aid themselves with supplementary bibliographical sources (e.g.dictionaries, encyclopaedias or technical papers) to more adequately determine which semes make up a given sememe.
Moreover, SV presents significant problems from the perspective of the philosophy of language.The lexicalisation of semes gives rise to entirely new semic sets, pertinent to the words used to lexicalise the original semes.Consequently, any attempt to maximise semic equivalence via the lexicalisation of the semes of an expression will only disperse its original semic configuration.Unfortunately, little can be done to overcome this linguistic paradox, for semantic analysis and replication cannot be carried out without words.Until telepathy supersedes language as the communicative tool par excellence, translators, linguists, and language users as a whole will have to cope with the imperfections of semantic analysis if they are to rely on the notion of meaning for any linguistic endeavour, regardless of whether it is for translation purposes, discourse analysis, or everyday communication.In practical terms, as long as CA and SV, from the semantic stance described in Section 2, prove efficient at sorting out problems on a superficial textual level -where linguistic paradoxes thrive without hindering effective communication-deeper philosophical dilemmas may be disregarded.
Be that as it may, translating via SV offers several advantages.Firstly, it is easy to implement.SV involves merely three sequential steps: (i) determination of the semes in an SL word through CA; (ii) lexicalisation of those semes in a SL; (iii) translation of the resulting lexical items into a TL. 10 This means that once translators acquaint themselves with the workings of SV, they may capitalise on its results at virtually no time-cost.Secondly, the semes determined via CA and then lexicalised through SV can be seen as a semantic sketch liable for syntactic reconfiguration.In other words, the semantic load of the lexicalised semes can be preserved while realising them in different grammatical units or word classes, as required by either the cadence or the thematic development of the TT.Thirdly, SV can also be used to overcome stylistic weaknesses of a phonological nature, since it can help the translator re-express the meaning of a cacophonic word by means of different-sounding lexical items stemming from the lexicalisation of the semes constituting the first item.Fourthly, SV is particularly effective in lowering the register of the translated text.This is so because lexicalised macrogeneric and specific semes tend to be much less formal and technical than the sememes from which they are derived.
Finally, SV may be of special interest to translation trainers and trainees.It would be interesting to carry out practical or experimental research on the topic in order to assess the results obtained through the application of SV.The argumentation presented in this paper, although liable to scholarly questioning as any other theoretical proposal, may find significant support if future empirical research shows that SV actually helps translators improve their praxis and substantiate their practical decisions._____________________________ 1 It should be noted that the idea of semantic replication implies that the words of any given language contain certain semantic features which will be necessarily present every time the words are read by the source readership.Moreover, it entails that such features can be transferred to a target language, where there would also be words comprising those exact aspects of meaning.Of course, this is not to say that all the semantic features in a word can be replicated in a target language.Instead, the underlying assumption is that specific meaning components are necessarily present in items belonging to different languages.On the one hand, those features are necessarily present because of the conventional nature of the linguistic sign, or, in other words, if certain features were not always present in a word, communication and intelligibility among the members of a speech community would be impossible or, at best, miraculous.On the other hand, the presence of those similar features in words belonging to different languages could be justified by highlighting the commonalities that all cultures have: if the ways in which different cultures verbalise their representations of experience were entirely dissimilar, translation would be impossible.Yet, these assumptions would nowadays reveal themselves largely debatable, since postmodernism would have it that meaning is not a stable property of signs themselves, but rather an effect of the different chains of signification which a text's signifiers trigger as they are recontextualised and resignified every time they are read or cited.In fact, from a deconstructive stance, there is no kernel or invariant of comparison across languages as might be posited by essentialist theories of meaning (Gentzler 1993).In terms of Kuhn (1962Kuhn ( , 1970Kuhn ( and 1977)), however, paradigms are incommensurable: the validity of an argumentation is to be established not by recourse to theoretical trends other than the one it stems from; contrariwise, it is to be assessed in terms of its own theoretical framework.If this paper relies on certain semantic notions which might be disputed from a post-modernist stance, it is because those very notions make it possible to support a useful practical rationale, which is the ultimate goal of the study at hand. 2 Katz and Fodor's theory was considerably revised and refined in Katz (1972). 3For an extended discussion on the opposition between essentialist vs. non-essentialist approaches to translation, see the scholarly debate held in Target (2000,2001). 4See Faber (1998) for more on language awareness and translation competence. 5This classification has been adapted from the one found in Louis Hébert's 'Interpretive Semantics ' (2006), in which only binary oppositions are acknowledged under the label of dimensions. 6Notice that there is no objective criterion for distinguishing domains or taxemes, which are to be established arbitrarily.Furthermore, in most cases, the distinction between those two categories proves blurry/vague. 7Such tendencies, however, are far from inflexible.since semes cannot be thought of or mentally conceived without recourse to a lexical representation of some sort. 9Notice that the macrogeneric seme [+UNKIND] does not necessarily imply the specific seme [+SCORNFUL].A person may be unkind without being scornful (for example, by insulting or ignoring any interlocutor). 10It must be noted that steps (i) and (ii), although distinguished here for theoretical purposes, would comprise a single step in actual practice.

Figure 1 :
Figure 1: A componential analysis of the sememe socks.

Figure 2 :
Figure 2: A componential analysis of the sememe mate.

Figure 3 :
Figure 3: A componential analysis of the sememe sneer.a sentence such as (7) The captain sneered at the rookie.could be paraphrased as (8) The captain smiled scornfully at the rookie.and subsequently translated as