Whether the natural language is motivated by non-linguistic factors, e.g., cognitive ones, has been controversial since antiquity. In Plato’s dialogue Cratylus, Socrates is asked by Hermogenes and Cratylus to act as an umpire on the truth or correctness of names, including proper names, common nouns, adjectives, or words in general. Cratylus’ position is known as naturalism. The names of the nightingale and the wolf do not (necessarily) refer to the sounds they produce. The louder and the more disturbing the horn of a police car, the more I am urged to give way. I can tell apart the cry of a wolf from the song of a nightingale. We similarly choose appropriate names that naturally fit the objects best. In contrast, Hermogenes’ conventionalism suggests that the police, medical emergency, and fire brigade cars have arbitrarily chosen a horn pattern so that everybody can recognize them running; they could have chosen their horning patterns alternatively.
The philosophical essence of Plato’s message is that both propositions are correct and wrong. There is arbitrariness but a motivated one. The issue spawned the theoretical question, what criteria determine the correct choice of name for any given object? Like an extreme linguistic conventionalist, Hermogenes holds that nothing but local or national convention defines which words we use to designate which objects. We could attach the same names to quite different objects and give single objects quite different names as long as the language users agree with the convention. Cratylus, like an extreme linguistic naturalist, argues that names cannot be arbitrarily chosen because names belong naturally to their specific objects. If you call something with any name other than its natural name, you fail to refer to it. For example, Cratylus tells Hermogenes, to the latter’s intense annoyance, that ‘Hermogenes’ is not his name.
Socrates starts by criticizing conventionalism and persuades Hermogenes that naturalism must be endorsed. A long central section starts in which Socrates explains his version of naturalism, appealing to etymologies of important philosophical words. As it turns out, those words have not been arbitrarily attached to their objects but are encoded descriptions. The argument seems to go Cratylus’ way until the final part of the dialogue. Socrates turns to Cratylus and shows him that his expectations as a naturalist are set impossibly high. In essence, names cannot aspire to be perfect encapsulations of their objects’ nature, and some conventional elements must be conceded
Aristotle pointed out that there is no natural relation between the sound of any word and the thing it refers to, taking thus a conventionalist’s position
Like Hermogenes, Aristotle, the Stoics, and those Roman Neo-Platonist philosophers, Saussure proposed that the relationship between a linguistic sign and the real-world thing it denotes is arbitrary. There is no direct connection between the form of the sign and the concept it represents
Ferdinand de Saussure is the founder of modern linguistics and semiotics
Nonetheless, despite the repeated praises for novelty, Saussure’s interpretation of the sign, in particular the verbal sign, as an indissoluble unity of two constituents – signifiant (signifier) and signifié (signified) – derives directly from the Stoic theory considering the sign (sēmeion) as an entity constituted of the signifier (sēmainon) and the signified (sēmainomenon). The latter is ‘perceptible’ (aisthēton), and the former is ‘intelligible’ (noēton) or ‘translatable’, to use a more linguistic designation
William of Ockham (also known as Occam, from Latin Gulielmus Occamus; c. 1287 – 1347) pioneered nominalism, the opposite of Plato’s realism. Some consider him the founder of modern epistemology. He strongly argued that only individuals exist, rather than supra-individual universals, essences, or forms. Universals (Plato’s forms) are the products of abstraction from individuals by the human mind and have no extra-mental existence
Incidentally, another significant contribution that Ockham made to modern science and intellectual culture, which I frequently recall herein, is efficient reasoning with the principle of parsimony in explanation and theory building. The principle of parsimony is also known as Occam’s razor. This maxim states that the most acceptable explanation of an occurrence, phenomenon, or event is the simplest, involving the fewest assumptions, possible causes, factors, variables, or changes without assuming hypothetical entities
The uttered word, the vocal opinion (sententia vocum), was another solution to the universals accepted in the early Middle Ages. Resuming Porphyry’s alternative, the first medieval philosophers (e.g., Roscelin of Compiègne or Rucelinus; c. 1050 – c. 1125) regarded genera and species (substance, animality, and humanity) either as things or as having no existence. By applying to this alternative Boethius’ terminology, nominalists thence derived either things (Latin res) or words (voces). To them, universals were the voices. This means, above all, (i) that universals are not things (only the individual exists) and (ii) that universals are merely words. For example, homo (the human species) is not an existent or mental entity but a name divisible into syllables, phonemes, consonants, and vowels. Hence, nominalists reduced the universals to the mere emission of sounds (flatus vocis). This Medieval concept is very close, if not identical, to Saussure’s sign with a signified (thing) and a signifier (word).
One of the reasons why Saussure’s hypothesis received so much attention and favour is that it came up at the right time, just after the establishment of Darwin’s theory of natural selection as a mechanism of biological evolution. Natural selection operates on random biological variants; therefore, language must consist of arbitrary (random) material to evolve. This material is the set of arbitrary names potentially given to a thing from which one arbitrarily prevails by chance. Language also evolves; therefore, it must change with the same device!
Comparative anatomy made Darwin’s theory of evolution possible; comparative linguistics should explain the evolution of languages. The analogies and similarities between languages call for a historical explanation and the postulation of an evolutionary tree. But linguistics seemed to have taken a Lamarckian rather than Darwinian theory as a model. Languages seemed to evolve purposefully. There was confusion between synchronic facts – how the variants of a word entered a grammatical system – and diachronic effects, the sound changes themselves. According to Darwin, any purposiveness in biological evolution does not lie in the changes (mutations, variants) themselves but wholly in natural selection, which is, in a sense, a synchronic process. New species develop from accidental, random mutations with no direction or purpose. Saussure saw Darwin’s principle as essential to a proper understanding of linguistics. Like genetic mutations, phonetic changes occur without a goal or purpose. Successful phonetic variants survive while others die out
Among the most fierce supporters of Neo-Darwinism in linguistics, Pinker and Bloom
Neo-Darwinism is a historical adaptation of the Darwinian theory to explain important scientific discoveries in 20th-century biology rather than a radically new theory of evolution. It incorporated, for example, the Mendelian notions and mechanisms of genetic mutation and recombination as a source of phenotypic variation
A link between language and genes was required for Neo-Darwinism to thrive in linguistics. Chomsky’s contributions to universal grammar theory came to establish that link. The general idea of this cognitive theory (considering language as a motivated, non-arbitrary system) is that a particular set of structural rules are innate to humans and independent of sensory experience. The human brain is structured to combine words that form grammatical sentences in a given language. This hypothetical structure is the language module or the language faculty. As I understand it, language differs from watch-making in that nobody knows how to make a watch from childhood. Still, practically everybody knows how to make sentences early in life because speaking is an innate brain faculty rather than being learned from the environment. The terms generative grammar, transformational grammar, or, most recently, minimalist program theory refer to more specific propositions within universal grammar theory.
Noam Chomsky – a prolific American linguist, philosopher, cognitive scientist, and historian – considers universal grammar as a byproduct of the evolution of our brain for recursive thinking, i.e., simply said, a capacity for putting things together and so producing new patterns from existing ones. In Mathematics, recursion is that for every number N, there is a number N+1. Although Chomsky rarely refers to genes or other genetic terms – he prefers mathematical and computer science paradigms – the supporters of universal grammar frequently do so. The link of language to genes is indirect and implicit. Language reflects our brain capabilities. The brain is a biological organ, and all biological organs develop according to underlying genetic information. Therefore, genes must ultimately structure and determine our language! Furthermore, if our genes determine the brain structure and its language faculty, then our language faculty and language must have been subject to mutation and natural selection and must have evolved in parallel with our genes. Since natural selection works on random variation to design systems that perform a function, language being a naturally selected system, is not random but fit to complete a task; therefore, motivated and not arbitrary.
However, everybody recognizes that language is not necessarily the direct object of selective pressure. Its evolution can be a byproduct of selection for a brain structure conferring other competitive advantages, e.g., recursive thinking. A famous analogy is the red colour of the blood. Nature did not select for the particular shade as it is usually thought to be doing with the colours of flowers visible to corresponding pollinators, but selected for haemoglobin as an efficient molecular carrier of oxygen to the tissues; haemoglobin just happened to be red. Another analogy may be the faculty of watch-making. Nature has selected brain and hand structures that enable us to handle or throw things, not directly for our ability to make watches with a competitive advantage (value for money). Watch evolution is not due to natural selection but socio-cultural selection (market rules). Darwin proposed the concept of natural selection as an analogy to artificial selection, i.e., husbandry methods for improving animal and plant traits.
There is always a risk of confusion between the human biological machinery serving language, i.e., the cognitive and motor traits required for speaking, listening, writing, reading, and thinking (the hardware), from one hand, and the language itself with its phonemes, morphemes, words, and higher structures (the software), from the other. The hardware is a set of biological traits, and, as such, it is subject to biological evolution, whether by natural selection or other mechanisms. The software does present variation, but this variation is not genetic, not inherited. It cannot be subject to ‘natural selection,’ as the term is defined in biological science today. We do not utter words randomly, letting the physical environment choose the fittest; we instead design our speech and writings and expect our social environment to respond. Language is mastered by training like muscles develop with work. One necessary consequence of the Neo-Darwinian theory of language is that hominids must have begun practising language hundreds of thousands, if not millions of years ago. By the time the human brain adapted its structure to using words according to universal grammar rules, words must have existed.
The acquired (or adaptive) immune system is another recursive biological system in which language may parallel is the acquired (or adaptive) immune system. Grossly, a handful of well-described hyper-variable genes continuously produce small amounts of proteins (antibodies) of practically every possible macromolecular form. Once a foreign protein (or entire organism) enters the host body, the immune system amplifies the production of those antibodies that fit best to parts of the invading structure to neutralize it. Natural immunity is polyclonal, i.e., several distinct (‘synonymous’) antibodies are selected and amplified for the same invader. Here we have genetic variation, recombination, and natural selection (differential proliferation and survival at the cellular level) operating before the immunologist’s eyes and within an organism’s lifetime. This type of immune system existed long before man – estimated to have appeared 500 million years ago – as it has been described in mammals and fish, birds, sea urchins, and perhaps even insects. Whereas the Darwinian (or Neo-Darwinian) theory may apply to the phylogeny of immune systems, immunity against a particular invader is a developmental biological phenomenon. It is not inherited and must be ‘learned’ and ‘memorized’ by each baby. What is inherited is the ability to develop immunity, which not only requires a handful of antibodies but also an array of structural and functional components of various cell types and organs, practically reflecting our entire genome.
There is a dialectical interplay of chance (variation, arbitrariness) and a purpose (motivation, selection, design) in all biological systems. Saussure has over-emphasized the arbitrariness of linguistic variation and overlooked the Darwinian concept of fitness (appropriateness, purpose). Biological variants are naturally selected for their fitness; some remain (survive) because they are more appropriate than others, the others are outcompeted and disappear in a given environment. If words and sounds evolve by this Darwinian mechanism, then the words we hear today are more appropriate than previously attested words for the objects they represent. The names of things within a language would not be arbitrary but the best possible descriptions in the context of a given culture; in a sense, motivated. The variation between languages would result from differential linguistic selection in different environments.
Nevertheless, the Darwinian theory has been accused of tautology: a surviving variant is necessarily fit since it survives and survives because it is fit (adapted). Tautological statements are unfalsifiable. Therefore, they do not make testable predictions and are not scientific
Saussure admits that the principle of arbitrariness applies only to the origin of some (most) words. Everything else in language, e.g., grammar and syntax, is motivated. He attenuated his fundamental principle of arbitrariness by distinguishing between the ‘radically’ and ‘relatively’ arbitrary elements of language. Once the random (arbitrary) words enter a linguistic system, they evolve within it, obeying phonetic, grammatical, and syntactic rules. For example, the formation of compound words and grammatical derivation are motivated phenomena and follow precise rules. Every subsequent change is, therefore, motivated and predictable. However, the rules are arbitrary because they vary from language to language.
The agreement with the Saussurean dogma of arbitrary signs was far from unanimous
Outside grammar, the structure of roots (Plato termed them stoikheia; elements) and indissociable or single-morpheme words (prōta onomata; primary names) reveal patterns inviting vigorous investigation of these questions. Even in primary names, there are patterns of letter usage correlating with meaning, e.g., the French enemi (enemy) and ami (friend); English father, mother, brother; ten, -teen, -ty; three, thirty, third; two, twelve, twenty, twi- and a twin; German zwei (two) and drei (three); Russian sem (seven) and vosem (eight); devjat (nine) and derjat (ten). In English, constellations of words having similar meanings share similar sounds, whatever the origin of such constellations may be: e.g., bash, mash, smash, crash, dash, lash, hash, rash, brash, clash, trash, plash, splash, and flash. The latter phenomenon is known as paronomasia or pons. This confrontation of phonetics and semiotics, irrespective of any etymological connection of the compared words, plays a considerable role in language life. Metaphor and metonymy assign the same signifier to a secondary signified, similar or contiguous to the primary signified
Similar to paronomasia is a phenomenon whereby phonetics is reminiscent of meaning. It is known as sound symbolism. For example, recurring phono-semantic pairings are found in the so-called phonaesthemes or phonesthemes
The same examples of echomimetic onomatopoeia and sound symbolism are mentioned in the literature repeatedly. Pimentel and colleagues reviewed the recent theoretical arguments, empirical studies, and experimental results accounting against arbitrariness
On closer scrutiny, onomatopoeia does not severely weaken the notion of arbitrariness either. The mimetic sounds differ significantly from one language to another. Saussure himself left some space for iconicity. But even iconicity can be arbitrary. He recognized, for example, non-random phonetic patterns such as the endings of active and passive verbs, the use of diphthongs starting with the phoneme /a/, compound words, and numerals, affixes, or the recurring counter-examples of echomimetic onomatopoeia. He admitted that arbitrariness is relative and that only part of the linguistic elements is radically (absolutely) arbitrary, presumably referring to the ‘hard’ roots of which the origin is unknown. What is arbitrary is the connection between the signifier and the signified in our minds. He attributes the origin of seemingly iconic ‘nuclei’ not to iconicity but chance. Those elements may subsequently establish themselves by some kind of linguistic selection. Saussure did not dismiss iconicity but limited its application to the relationship between the signs and their referents, which falls outside linguistics as he defined it. Hence iconicity does not contradict arbitrariness. It instead applies to the relationship between signifier and signified within the linguistic sign
Moreover, even if statistically significant tendencies for iconicity exist – for example, gl- may appear in words denoting light or shiny objects more frequently than expected by chance alone – this kind of regularity fits Saussure’s framework. If an ancient root meaning light contained some combination of letters, all the cognate derivatives of that root might preserve those letters. The GL happened to be the combination arbitrarily chosen in English. The choice of these letters in the ancient root remains arbitrary. Different phonaesthemes may arbitrarily occur in different languages. To falsify a hypothesis, one needs to demonstrate facts that are incompatible with that hypothesis, not merely facts that are compatible with another explanation. To accept phonaesthemes as falsifiers of arbitrariness, one would need to show that they are incompatible with this hypothesis rather than that they are compatible with a suggestion of non-arbitrariness
A theoretical and practical shortcoming of the arbitrariness hypothesis is its predicting impotence. It is impossible to infer an unknown word’s meaning from its sound
Saussure’s arbitrariness is ill-defined at the start. For this reason, any motivated or iconic words can be accommodated in his theory and ‘explained’ without difficulty. In mathematics, arbitrary means any, out of all, that is possible, undetermined, not assigned a specific value. As an epistemological term regarding actions or ideas, arbitrary means subjective, unreasonable, irrational, unscientific, and unpredictable. Randomness and arbitrariness are tightly related to unpredictability. Something based on individual discretion or judgment rather than on an objective distinction is not a scientific object, despite the self-proclamation of Saussurean linguistics as a science. Building a bridge is a scientific problem; where to build a bridge is not a scientific question. We cannot predict a random or arbitrary outcome. Neither can we predict the meaning of an arbitrarily (randomly) formed term. In essence, a random natural event or an arbitrary human thought is not something with no causes (no motivation) but something too complex to handle. Once underlying patterns emerge, the object becomes predictable, hence scientific, but it is, by definition, no longer ‘arbitrary’ or random.
Like all rational human behaviour, a language may simultaneously be arbitrary, motivated, and conventional. How an architect will arrange the rooms of a house in a given plot is an arbitrary decision, but it is motivated by a rationale, and it is restricted by building laws and finance; no architect works at random. Saussure used the term arbitrary as the opposite of iconic or motivated: not representative or symbolic; not having the characteristics of, or relating to, an icon; not famously and distinctively representative of its type; not representing something; not based on a rationale; therefore random. As absurd as it may sound, arbitrariness is to draw five-letter pieces from a well-shuffled scrabble bag when we want to give a five-letter name to something. Words do not seem to have formed this way. Saussure based his assumption that different languages synchronically have different signifiers for the same signified. He was much less concerned with diachronic changes and original roots, which, he admitted, might have been motivated
Following a purely theoretical approach, Horst Ruthrof argues that Saussure’s notion of arbitrariness is complex. It comprises the arbitrariness of the signified, the signifier, the relation signifier-signified, and the sign as a whole. He claims that language is only a tiny part, a refinement, of the human semiosis, which mainly consists of motivated, nonverbal signs. From his perspective, language is parasitic on nonverbal signification, particularly in body language. If nonverbal signs are motivated, and language cannot function without nonverbal semiosis, then the linguistic sign cannot be arbitrary in the Saussurean sense. He, therefore, proposes a redefinition of the linguistic sign as a hybrid entity made up of an arbitrary signifier and a motivated signified. This distinction would allow for natural language to straddle the domains of formal and sufficiently iconic sign systems without being identifiable with either
Our access to linguistic facts is rapidly changing. Recent research reveals that forms of non-arbitrariness are more widespread than previously assumed. Such findings suggest a more textured vocabulary structure. Arbitrariness is complemented by systematicity when statistical regularities in forms predict function. It is also complemented with iconicity when word form aspects resemble meaning in various languages. The latter cases are called ideophones
To reverse-engineer ancient names and understand the meaning of myths as encoded in them by their authors
The strategy followed herein to falsify the arbitrariness hypothesis about the link between the signifier (word) and the signified (think) is to demonstrate the following: (i) The landmark examples of arbitrarily formed words such as the English tree, dog, cat, etc. are iconic. (ii) Many other inconspicuous hardcore root morphemes are also iconic. (iii) Iconic primary words and roots constitute a significant proportion, if not the majority, of the vocabulary, at least in English and Greek. (iv) Double phonemes, like /ps/ of the Greek letter Psi, contain composite sememes. (v) Aspirated consonants and digraphs such as Th, Ch, Sh, Ph, etc., as well as the double letters, like MM in the English hammer, denote duality, plurality, intensity, repetition, or quantity. (vi) Inversion of a sequence of phonemes or letters is associated with reversal of meaning (antonymy). The former three observations diminish the field of application of Saussure’s concept without necessarily falsifying it. They predict, nevertheless, that roots which remain unexplained are also iconic. The latter three principles are utterly incompatible with the assumption of arbitrariness. If validated independently, iconicity will give us new perspectives on studying ancient languages and mythology.
References
