[I have noticed that some people fight shy of long essays -- doubtless an effect of the syndrome lamented here. Accordingly, as an experiment, we shall intitially put up just a bit -- a “stub”, in Wikipedia’s terminology; or, as Professor Malkiel loved to say, a “torso” -- adding to it as the days go by,
as the moon rises and the sun sets,
as the leaves fly off the calendar in the winds of Time ...]
The ontology, then, of “linguistics”: and not, note, of “language”. Similarly, we may speak of the Ontology of Psychology (and not of the psyche), the Ontology of Geology (and not of the Earth). All That Is, is what it is; “I am that I am”. But for purposes of this or that variety of study, we structure things. Such structures are constrained by what is Out There, but are not straightforwardly or uniquely determined.)
So, first of all: ‘What is linguistics?’
(We cannot pose this tiresome question, save in squotes, just as we did for ‘What is Mathematics?’)
Two candidates present themselves:
(I) Linguistics is the study of language.
(II) Linguistics is the study of languages.
In many modern perspectives, these are distinct. And each, in its own way, is problematic.
(I) Already with Saussure if not before, the nature of the pre-theoreticaly notion ‘language’ (langage) was split, for precision, into langue and parole. Subsequently, this basic bifurcation acquired theoretical heft with, in one corner (in the red trunks) those championing “I-language” and the innate “Language Acquisition Device”, versus (in the blue trunks) corpus-mavens, connectionists, frequentists, “usage-based” grammarians, and other sundry nominalists. For a riposte to the latter, confer Frederick Newmeyer, whose tautologically-titled essay “Grammar is Grammar and Usage is Usage” (in Language, 2003), makes many useful points.
(II) This formulation is less problematic, as being less ambitious, and more traditional. Indeed we may say: Philology is to languages, as linguistics (in the contemporary sense) is to language-tout-court. But then we are faced with the question, what is “a language” -- as opposed to a different one, or a dialect, or some other semantic signaling system. And there we meet disagreements once again.
~ Recommendation posthume ~
“Si j’étais encore en vie, et que je désirais un bon whodunnit,
(Je suis Ferdinand de Saussure, et j’ai approuvé ce message)
This antinomy of language-per-se versus languages, has become acute since Chomsky, who has little interest in the endless gabble of actual tongues. If this seems an extreme position, it is nonetheless exactly parallel to that of physicists, who seek general physical laws, rather than endless descriptions of individual objects falling or rolling or spinning or colliding or what have you.
Chomsky bitingly writes:
The grammar is a function-in-intension … the language is epiphenomenal. Its ontological status is the same as that of a set of pairs of expressions that rhyme.
-- Noam Chomsky, Rules and Representations, p. 83
Here he is using language in the sense of ‘parole’ and not ‘langue’. But, epigrammatically, it is startling to see language as of but peripheral interest to linguistics.
Marveling at the philosopher’s opacity, two linguists write:
In a recent article, Ryle even claims that sentences are not part of language, but only of speech.
--Jerry Fodor & Jerrold Katz, eds., introduction to The Structure of Language (1964), p. 11
Yet Chomsky would later also demote much of what was traditionally thought of as language, to a second-class status as “E-language”.
The units of social life are far less clearly defined than those of language … Linguists are fortunate in possessing a domain whose units are at least relatively self-defining and isolable.
-- Ernest Gellner, Contemporary Thought and Politics (1978), p. 82
The minimal unit of spoken language is not the sentence, but the utterance or intonation unit.
-- William Bright, ed. International Encyclopedia of Linguistics (1992), vol. IV p. 11
Schuchardt enquires, “What is dialect?” -- the neogrammarians having used the term in formulating their theories -- and shows that it is an abstract notion, with no real existence.
-- I. Iordan & John Orr, An Introduction to Romance Linguistics (1937), p. 32
With similar vigour, Schuchardt opposes the notion of ‘linguistic periods’ .. Just as there are no fixed boundaries between different vernaculars, so the chronological boundaries between successive periods of a language are purely fictions of our minds.
-- I. Iordan & John Orr, An Introduction to Romance Linguistics (1937), p. 33
Die Gesamtsprache ist etwas Abstraktes, ebenso wie die Gesamtseele gegenüber der Individualseele.
-- Leo Spitzer, ed., Hugo Schuchardt-Brevier (1921; 2nd edn. 1928), p. 386
The term language is … a relatively nontechnical term. If we wish to be more rigorous .. we have to employ other terminology. We shall use variety as a neutral term to apply to any particular kind of language which we wish … to consider as a single entity.
--J.K. Chambers & Peter Trudgill, Dialectology (1980), p. 5
(Ontologically/methodologically, this picture is not particularly reassuring.)
Structural linguistics has a favored suffix to denote status as a unit in the ontology or theoretical apparatus: -eme. Familiar (though embattled) are the phoneme and morpheme; lexeme (roughly: dictionary headword, so that eat and ate are part of one ‘lexeme’ even if you might call them different ‘words’ in one of the senses of that pre-theoretical term). Hjelmslev proposed seme (or sememe), as the atomic unit of sense (the ‘semantic atom’), though these do virtually no work; and, getting into the borderline-silly spirit of the thing, from outside of linguistics Richard Dawkins coined meme.
The ontological status of ‘meanings’:
Quine is quite correct in protesting that meanings are not entities.
-- Harold Lee, “Discourse and Event”, in: Hahn & Schilpp, eds., The Philosophy of W. V. Quine (1986), p.
One might have thought that being an ‘entity’ would be rather a low bar.
(I recall one smitten young man, back in Berkeley in the 1970s, pleading with her whom he adored, “What do I mean to you?” -- And she replied, wrinkling her young brow and thinking hard, “Well, … you’re a person ….” )
It goes even lower:
Though we sometimes need to conceive meanings timelessly, we do not therefore need to conceive them as subsistent entitities.
-- Jonathan Cohen, The Diversity of Meaning (1963), p. 161
(“Subsistent”: cf. Meinong. Actually, “subsisting” is a very low bar, lower than “existing” entities.