1.1 Approaches to word formation
As a branch of linguistic science, word formation is concerned with analysing and understanding the mechanisms by which the lexis is created and renewed. These mechanisms are mainly morphological, involving the combination of words and their subunits in various ways, although word creation may also involve other procedures such as borrowing words from other languages, the formation of new terms through the combining of initial letters of the names of institutions, generally known as acronymy, or the subtraction of units from words in processes known as back formation or clipping. The data with which word formation is concerned is to be found not only in the existing lexis of the language as recorded in the dictionary, but also in the neologistic terminology of science, technology, commerce, the mass media, in the creative language of modem literature, and in the colloquial and innovatory language of contemporary speech.
In handling these sources of data, the approach may be diachronic, looking back to the etymological sources of word forming procedures and examining the dominant morphological patterns of the past; or it may be synchronic, referring to present-day tendencies which will determine the features of the vocabulary in the immediate future. It may be illuminating to combine these approaches as Marchand (1969), for example, has done to a certain extent in his exhaustive account of English word formation. Whether diachronic, synchronic, or both combined, a variety of criteria has to be applied in attempting to account adequately for the many regularities as well as the real or apparent irregularities which characterise the procedures of word formation. Etymological analysis can show which patterns are due to Greek or Latin sources and how word forming units evolved in meaning and function in the evolution of Castilian. Morphological criteria will be applied in describing the actual permissible combinations of units, and these in turn are invariably governed by the dominant sound features of the language requiring the application of the principles of phonology. Semantic considerations have to be taken into account in explaining relationships between word components and the result of their combination with regard to meaning. Since the final result of these procedures is a new term which will require immediate or eventual entry in the dictionary, then lexicographical implications ensue. So word formation involves aspects of all the major divisions of linguistic analysis, making it a particularly complex area of study and bringing together many critical problems of present-day linguistic theory.
Different schools of linguistics, while exploiting all such criteria, have tended to favour one or another at particular times. In the pretwentieth century diachronic focus, attention was concentrated on the striking transformations of the lexis over the centuries, and in particular on the way in which the vocabulary seemed to be structured in historical layers through successive periods of different ethnic influences.1 Thus the constituent parts of the vocabulary of Spanish were perceived as being Latin, Greek, and Arabic words, as in the Arabic series in a/- (<algarroba, alguacil, alfiler, etc.), or word parts as in the latin suffix -tas (sinceritas, Jutilitas, gravitast etc.) which yielded the long Spanish abstract noun of quality series in -dad (sinceridad, futilidad, gravedad, etc.). These historical criteria were reflected in the dictionary where the current forms of the entries were explained by their etymological origins, as exemplified in the Vox dictionary (1964):
Although three of these items show compositional morphological structure the emphasis in the entry is on the etymology rather than on the structure. In the same way, in specific studies of word formation applied to individual languages, great attention is always paid to etymology, as for example in Tekavčič’s account of Italian derivational morphology, where the Latin and Greek sources of the components are given priority in the descriptions (Tekavčč: 1972). This emphasis is seen in the strongly diachronic approach of traditional studies of Spanish word formation as well as in the treatment of derivational morphology in traditional grammars. So, for example, the comprehensive pedagogical grammar of Ragucci (1963) which distinguishes itself by the close attention given to an aspect of language often skirted in grammars, deals with it in a chapter headed:
Breves nociones de Etimología. -Formatión de las palabras. -Derivatión, compositión y parasíntesis. - Palabras primitivas y derivadas; simples, compuestas y parasintéticas. -Análisis etimolégico.
In this approach, etymology and word formation are thus regarded as one, with suffixes and prefixes classified on an etymological basis (castellanos, latinos, griegos). Similarly, the distinction of Latin, Greek, and Castilian is basic to the Spanish Academy’s approach, as shown in the Gramática de la Lengua Española of 1931.
In the early twentieth century, the advent of Saussurian linguistics marked a change in emphasis away from the historical account of language to a synchronic descriptive approach, studying the systems and rules of the internal mechanisms of language, independently of the historical or ethno-cultural environment. Here, however, word formation was not in the forefront of interest; on the one hand it combined diachrony and synchrony which the new linguists were keen to keep separate, and on the other it was concerned with productive procedures going beyond synchrony and looking towards the future state of the language by way of lexical change and innovation. In the post-Saussurian period, interest concentrated not on the word but on the minimal segments of speech as represented by the morpheme and the phoneme and analysed without close regard to their combination into larger units, and the very status of the word was questioned as a useful unit of analysis.2
Just as the advent of Chomskian linguistics from the late 1950s marked a dramatic innovation in general linguistic theory, it eventually did so also in the treatment of word formation. Although initially transformational-generative grammar was concerned with syntax in its attempts to explain the creativity and competence of the native speaker in producing and understanding an infinite number of new sentences, the problems posed by the plethora of structure models which emerged led to a new interest in the word, especially in its function as the lexical insertion component in deep structure and its syntactic relationship with the rest of the sentence. At the same time, by stressing the creative rather than the prescriptive aspect of the grammar, transformationalism could no more overlook the ability of the native speaker to speak and understand new words than it could the ability to construct and understand new sentences. As well as an innate grammatical faculty, the native speaker was perceived to be endowed with an inherent lexical competence, which is the basis of the lexicon and of neologistic terminology not yet recorded in the dictionary. In this way, from being marginal, word formation has come to play a central role in general linguistic theory, and under the label of ‘derivational’ or ‘lexical’ morphology become the raw material of a wealth of modern theories evolving from transformational-generativism.3
In the later development of transformational-generative grammar, word formation ceased to be isolated from phrase structure and sentence formation. Indeed, ‘the development of transformational or transformational-generative grammar from its beginnings up to the present can be seen, among other ways, as a progressive refinement of the structure of the lexical component’ (Scalise 1984: 1). Following the tenets of post Chomskian syntactic theory, the procedures involved in forming words were taken as being analogous to those involved in forming new sentences. The form of complex words was seen as itself containing syntactic structure, the derivative or compound being no more than a surface representation of this, a sort of graphic shortcut. Stemming from this, linguists enthusiastically applied transformational analysis to the lexis in an attempt to explain word formation on some logical basis. An example of this is the prolific work of Guilbert in French, both in the study of individual lexical subject areas and in the dictionary as a whole.4 In Spanish this might be exemplified as follows, where a) = derivative lexeme and b) = underlying syntactic structure:
In terms of transformational-generativism these represent surface realisations of transpositions from verb to noun structures. This approach may be specious in seeing everything in terms of deep structure and tending to be over-preoccupied with the verb and noun phrase relationship which fascinated the post Chomskian linguists. It tends, for example, to overlook other types of syntactic links in word formation, such ...