Words have conventional meanings, but do they also have their own meanings, apart from the meanings that humans assign to them? For an object-oriented linguistics, a word is its own thing, distinct from any references or any speaker or listener that encounters it. This would suggest that a word has its own meaning, its own style of being, its own virtual capacities. This opens up old questions about how much meaning is given in the thing itself and how much it is determined by the word’s references and relations to the conventions and contingencies of speakers and listeners?
Plato’s Cratylus revolves around the question of whether names are given by nature, which is the position of Cratylus, or by conventional agreement, which is the position of Hermogenes (383a). Through a series of questions addressed to Hermogenes, Socrates arrives at the conclusion that a name is given by nature if it imitates the essence (and not the accidental qualities) of the thing it is naming. “Well, then, if anyone could imitate this essential nature of each thing by means of letters and syllables, he would show what each thing really is, would he not?” (423e). Furthermore, even individual sounds and syllables have the potential to symbolize the essence of a thing. However, Socrates is aware of the difficulty of taking language to pieces to see if and where this symbolism is present: we only have our opinions upon which to initiate such a task (425b-c). Investigations attempted subsequent to Plato’s Cratylus have espoused many more opinions on the question of the relation between the sounds of names and that which they name: Stoics, Medieval nominalists, and modern linguists have all taken up this question of the relationship between signifying sounds and signified meanings. A standard position held by many scholars today is that a signifying sound has an arbitrary relation to that which it signifies. Even many textbooks in linguistic theory dogmatically presuppose the arbitrariness of the sign without even hinting at alternative interpretations of the sound/meaning relationship, except for brief mention of onomatopoeia, which is usually shown to still admit of arbitrariness.
Despite the prevalence of the theory of l’arbitraire du signe in modern linguistics, there have been numerous investigations that demonstrate a much more intimate relation between sound and meaning—a relation called “sound symbolism” and “phonetic symbolism” by Edward Sapir, and called “phonosemantics” in some recent studies by Margaret Magnus and John Lawler among others. This phonosemantics, although present in all language, is immediately apparent when a sign is not reduced to its relations to human speakers/listeners or to any autonomous signified, but when it is understood as a sign, as having its own structure and meaning, functioning poetically (in Jakobson’s sense of poeticity). When understood poetically, equations of sounds and meanings appear in sequence(s) on the axis of combination, where “-ing” always means “-ing” and “str-” always means “str-”, and within every phoneme and morpheme are differential relations folding together the whole phonology and semantics of the language.
These sounded meanings of messages themselves are virtual capacities that become manifest in references to a signified or in relations to speakers and listeners. What are the phonosemantic capacities of messages themselves, and how do they become manifest in the relations and references of the messages? Consider a couple examples:
In “A Study in Phonetic Symbolism” (Selected Writings, 1949), Sapir attempts to demonstrate that, despite the arbitrary allocations of sounds in conventional fields of reference, “symbolisms tend to work themselves out in vocalic and consonantal contrasts and scales” (p. 62). Sapir discusses an experiment wherein subjects were asked to discern the relative sizes of objects that were designated by artificially constructed words. Where the word mal was defined as “table,” the majority of subjects defined the word mil as a “smaller table.” Thus, the /a/ in mal was perceived to symbolize a larger table than the /I/ in mil. This perception was not based on an established convention, because the words were artificially constructed.
Sapir hypothesizes that this symbolism may be due to some combination of the acoustics of /a/ and /I/ and the kinesthetic experience of /a/ and /I/ (p. 69). Acoustically, /a/ has both more volume and more duration than does /I/. Kinesthetically, /a/ is articulated with the tongue in a much lower position than when articulating /I/. Thus, /a/ both sounds larger and feels larger than /I/. From this, Sapir observes “that the symbolic discriminations run encouragingly parallel to the objective ones based on phonetic consideration” (p. 68). In other words, when a signifying sound occupies more space and time, it symbolizes a larger signified than a signifying sound occupying less space and time. To put it simply, bigger sounds have bigger meanings, which can become manifest in references to bigger things. The big/small opposition is given with a vocalic opposition, and thus can accommodate big/small oppositions that appear in particular contexts.
This does not mean that the meaning of /a/ is largeness. That is one manifestation of the virtual capacities of /a/. Situated in another language, the same phoneme could manifest very different capacities. Consider the words of night and day in French and Czech, particularly in light of their vowels (which are grave and acute, roughly corresponding to Sapir’s /a/ and /I/, respectively). The Czech word for day, “den,” symbolizes the brightness of day (a chromatic sensation) with an acute vowel, whereas the Czech word for night, “noc,” symbolizes the darkness of night with a grave vowel. However, the French “jour” (day) has the grave vowel in relation to the more acute “nuit” (night). This apparent contradiction of symbolisms is really just different manifestation of the virtual capacities of grave and acute vowels. Brightness is but one capacity. The French language activates not the darkness but the intensity of the grave vowel, such that jour can refer to the work and movement that fill the day in contrast to the relaxed and mild motion of nuit. In sum, the capacities of phonemes (e.g., grave and acute vowels) can become manifest in multiple ways, and those manifestations can enter into relation with manifestations of the various capacities of things (e.g., brightness and intensity of days and nights).
These are just some initial thoughts on object-oriented linguistics, the main point being that linguistic signs are phonosemantic things, which have virtual capacities (poeticity) as well as manifestations (i.e., references to signifieds and relations to speakers/listeners). I’ve shown that a phoneme is a phonosemantic thing, but linguistic thinghood is not only at the phonological scale. Morphemes, words, and sentences are also things. I’m not sure if individual phonemic features are things or just parts or qualities of things (features like labial or dental, nasal or oral, stop or fricative). A question I want to think about more is markedness, specifically in showing how a linguistic thing changes, develops, and enters into new relations.