The semasiological description of the significative system of the language is based on purely structural concepts. It may therefore also be called ‘structure-based description’.

Main subdivision

The semasiological description is introduced by a chapter on structural classes, relations, and processes. It classifies the categories of words and formatives and enumerates the elementary concepts by which constructions are formed. The ensuing top-level subdivision is by the principal categories of syntactic constructions, established as distribution classes. Its structure is as follows:

particle - adjectival - nominal - adpositional - adverbial - verbal - clause - sentence.

The idea behind this order is to treat first the building blocks and last the top-level category of grammatical description. A language may, of course, have different principal syntactic categories.

By definition, the category ‘particle’ does not project onto a complex category. This chapter therefore only classifies the particles by their distribution.

In principle, and in analogy to other chapters, the complex clause would be dealt with in the chapter on the clause. In many languages, however, this is bound to be a complex and long chapter; so it may be taken out and positioned between the chapters on the clause and on the sentence.

Hierarchical subdivision

Within each of the above principal categories except particles and sentences, subdivision is by constructions headed by an item of the respective category, ordered by increasing complexity, roughly upwards the hierarchy of complexity levels in grammar:

Specifically, each of the chapters from the adjectival up to the verbal construction is, in principle, subdivided as follows:

pro-form - stem formation - inflection - complements - modifiers - negation - coordination.

A subsection on phraseologisms of the respective syntactic category is appended to this sequence.

Each section is devoted to a particular construction. Inside each of the above principal construction types, subtypes are generated by their internal composition. In the simplest case, the construction has a binary structure, consisting of a head and a dependent. The head is the item that determines the syntactic category at the highest hierarchical level. The dependent may be formed by elements of different syntactic categories that bear different relations to the head. An important distinction here is between complements and modifiers. Specifically in nominal constructions, there is a third function, viz. the determiner. Components bearing a particular dependency relation to a head may differ by their internal composition. For instance, an attribute may be constituted by an adjective or by an adpositional phrase. These are criteria for further subdivision.

The syntagmatic relations obtaining in a particular construction are visualized in the form of a construction schema. It has a place in a taxonomy and in a meronomy. These hierarchical relations are made explicit with each construction schema, so inheritance relations follow automatically.

In a binary construction, the head or the dependent may belong to a closed class. Its paradigmatic relations – i.e., the set of items of the same class – are visualized in the form of a table.

The criteria of classification are strictly distributional all the way down. For instance, the clause has a component that functions as adjunct. At a lower level, adjuncts are subdivided by the category of the item forming them, like adverbials, adpositional phrases, gerunds etc. Down to this point of the subdivision, no semantic criteria are involved. For instance, if there is a section on temporal adverbials, it is below the level of adjuncts constituted by an adverbial and, expectably, at the lowest level of subdivision of the semasiological grammar. The subclassification ends at this point; considerations of meaning and use structure the onomasiological description.

The description of a syntagma of category C is goal-oriented. This is to say, it is devoted to the problem: What are the possible ways of forming a syntagma of category C; which variants of the internal structure of C exist? Conversely, the question is not: What are the possible uses of a syntagma of category C? For instance, the chapter devoted to verbal constructions has a section on stem formation. The question there is not: which kinds of derivation are available for a verbal base? Instead, the question is: what are the derivational operations to form a stem of the category ‘verb’. The possible uses of C, i.e. the higher-level constructions of which it is a component, may be enumerated at the end of the section devoted to C, with references to the sections of the next-higher complexity level devoted to the more inclusive target construction.

An important feature of this arrangement is that there is no division between morphology and syntax. Thus, problems such as the distinction between compounds and phrasal compounds or the alternative of assigning clitics and periphrastic constructions to morphology or syntax do not arise.

The following table summarizes schematically the full structural description of a construction.

The semasiological description includes an alphabetical dictionary. It contains at least all the words and grammatical morphemes of the language which occur in the text, the examples and the specimen corpus. The microstructure of an entry consists minimally of the lemma, its word class and the set of senses, ordered first by syntactic, then by semantic criteria and including selection restrictions. This section is included for practical convenience. From a systematic point of view, words would be classified by distributional and morphological criteria and be treated inside each of the preceding chapters.

Some aspects

Item focus

The approach “we have a little word here; let's see what one can do with it” is familiar from school-teaching and hundreds of philological dissertations of past centuries.¹ It is necessary for lexicography; but it does not correspond to any known linguistic reality, in particular not to the point of view of the speaker/hearer or the learner; and it is inappropriate in grammar. A systematic structural description is not item-focused. It first presents the syntagmatic and paradigmatic structure and only then arrives at a given position in the syntagma and in the paradigm, occupied by a particular item. An example is in the section on explicitness of linking.

Basic meaning

The semantic and functional analysis of items and constructions of a language the analyst does not speak well must, of course, proceed with caution. The first task is not to look for a translation equivalent in more familiar languages, but to bring out the peculiarities of the phenomenon in this particular language by a sober semasiological analysis.

Once this is done, however, the reader has a right to know what the closest translation equivalents in more familiar languages are. Semantics relies strongly on the hermeneutic method, and the understanding the reader may achieve after working through a page of detached structural analysis can often be reached much more readily by a single translation equivalent in his native language. American structuralists of the mid-twentieth century have typically shunned away from this kind of linguistic explanation. However, as long as it is not used to replace an analysis, but just offered the reader as a crib, there is nothing against it. On the contrary, withholding this kind of understanding from the reader may mislead him, as it violates Grice's maxims of manner #6 and #8. Consequently, if the author keeps silent about what appears to be the most straightforward way of conceiving the matter, the reader may infer that the expression does not have the meaning that he thinks he has understood.

¹ Doctorandus, Iustus 1865, “De functionibus coniunctionis ut in operibus C. Iulii Caesaris”. Institutum Philologicum Collegii Oxfordiensis: Dissertatio submissa pro obtentione tituli doctoris philosophiae.