What Kind of Thing Is a Phoneme? Pāṇini, Saussure, and Jakobson
Is a phoneme a concrete sound or an abstract entity? The question seems straightforward, but it becomes difficult as soon as we ask what we mean by phoneme. The same phoneme can be realized in somewhat different ways, while a relatively small difference between two sounds may be enough to distinguish two phonemes within a particular language. What, then, is the thing we are calling a phoneme?
Part of the difficulty is that phoneme does not designate exactly the same object in every linguistic theory. The term has a history, and its meaning changes with the theoretical system in which it occurs. This is particularly important from a Saussurean perspective. Just as the value of a linguistic sign depends on its relations to other signs, a theoretical term derives much of its value from the concepts and distinctions surrounding it. We might therefore think of a linguistic theory, for the purposes of this discussion, as a kind of “mini-langue”: a structured system in which concepts acquire their value through their relations to other concepts. Phoneme in one such system need not have exactly the same value as phoneme in another.
The history of phonology offers a revealing case. The Indian grammatical tradition associated with Pāṇini developed an exceptionally precise classification of Sanskrit speech sounds long before the emergence of modern phonological terminology. Patañjali, in his discussion of sphoṭa, also confronted, in a very different conceptual framework, the problem of linguistic identity and variable realization. Centuries later, Saussure approached the problem from the perspective of historical reconstruction and, eventually, of the linguistic system itself. Jakobson would later develop a theory in which phonological oppositions could be analyzed through distinctive features.
These traditions should not be arranged as though one simply discovered what the others had missed. Pāṇini, Saussure, and Jakobson were working within different intellectual frameworks and asking different questions. Their importance for the present discussion lies precisely in the differences among their approaches.
The purpose of this article is therefore not to decide which of them was “right,” nor to construct a simple historical line from Pāṇini to Saussure to Jakobson. Instead, we will follow a problem: what makes a speech sound into a linguistic entity? Is such a unit defined by the physical properties of the sounds associated with it, by its relations within a system, or by some combination of the two?
The student's question—Is the phoneme concrete or abstract?—will serve as our starting point. But as we follow the history of the problem, the question itself will become more difficult.
1. Before the Phoneme: What Did Pāṇini Analyze?
The word phoneme belongs to the vocabulary of modern linguistics, so it would be misleading to begin by saying that Pāṇini “discovered the phoneme.” What the Indian grammatical tradition achieved was something historically more specific and, in certain respects, surprisingly close to later phonological analysis: it developed a systematic way of organizing the sounds of Sanskrit according to recurring properties and relationships.
As Georges Mounin emphasizes in Histoire de la linguistique, the Indian grammarians did not merely distinguish vowels from consonants. Their analysis classified speech sounds according to manner and place of articulation, taking into account both the location of articulation (sthāna) and the articulating organ (karaṇa). The classification also distinguished properties such as voicing, nasality, and aspiration. Mounin calls attention to the remarkable organization of the Sanskrit consonantal system into correlated series in which voiceless and voiced sounds, their aspirated counterparts, and nasals are systematically related.
The important point is not simply that the Indian grammarians knew a great number of phonetic facts. It is that these facts were organized systematically. A sound was not treated as an isolated event but as a member of a structured series. What distinguishes one sound from another could be described through dimensions that recur across the system.
This makes it tempting to call the Pāṇinian analysis “concrete,” especially because it is attentive to the articulatory and phonetic properties of speech. But that description can be misleading if it suggests a mere inventory of physical events. The categories abstract from individual utterances and organize sounds into recurring classes. The relevant object is therefore not a single pronunciation but a systematically defined sound class.
The distinction becomes more interesting with Patañjali, Pāṇini's great commentator. Mounin presents the theory of sphoṭa as addressing, in its own way, a problem that will remain central to later phonology: how can something remain linguistically identifiable when its concrete pronunciation varies? One speaker may pronounce an utterance rapidly, another more slowly, and another with somewhat different phonetic characteristics. The realization changes, but linguistic identity is not thereby destroyed. Mounin describes sphoṭa as an invariant substratum distinct from the individual phonetic realizations through which it is manifested.
It would be anachronistic, however, to identify sphoṭa straightforwardly with the modern phoneme. The two concepts belong to different theoretical traditions and perform different functions. The value of the comparison lies elsewhere. Patañjali makes visible a distinction between variable phonetic realization and linguistic identity. The same problem will return, in a very different conceptual vocabulary, in modern phonology.
The Indian grammatical tradition therefore gives us an early and sophisticated way of separating two questions that are easily confused: what properties do speech sounds have, and how are those sounds organized into stable linguistic categories? Its analysis does not provide a ready-made theory of the phoneme, but it demonstrates that the relation between physical realization and linguistic organization can be investigated systematically without reducing one to the other.
This is one reason the Indian grammatical tradition is important for the history of phonology. Mounin stresses that its analysis of sound units reached a level of sophistication that European linguistics would attain only much later, while also noting that the achievement remained largely without influence in the Western history of linguistics because it was not properly known or assimilated. Its historical significance is therefore not that Pāṇini supplied a theory later borrowed by European linguists. Rather, the Indian tradition shows that highly systematic analysis of speech sounds existed independently of the conceptual vocabulary that would later develop around the phoneme.
For our purposes, Pāṇini and Patañjali leave us with two related problems. The first concerns the systematic classification of speech sounds and their properties. The second concerns the persistence of linguistic identity across variable realizations. Neither problem can be settled simply by pointing to a particular physical sound.
The modern history of the phoneme begins when these questions are taken up again under very different historical and theoretical conditions.
2. The Nineteenth Century: The Sound Is Rediscovered, but the Problem Changes
The rediscovery of Sanskrit in European scholarship at the end of the eighteenth and beginning of the nineteenth century created the conditions for a new science of language. Yet the transition was not as direct as we might expect. Europe encountered a linguistic tradition whose analysis of speech sounds was extraordinarily sophisticated, but the first generation of comparative linguists did not immediately adopt that analysis.
Mounin emphasizes the paradox. The importance of Sanskrit for European linguistics between roughly 1786 and 1816 is undeniable, but its influence on phonetic observation was initially limited. Early comparatists continued to reason to a considerable extent in terms of letters rather than sounds. The discovery of Sanskrit therefore did not simply transfer an existing Indian analysis into European linguistics. Instead, it entered an intellectual environment increasingly concerned with historical comparison and systematic relationships among languages.
The gradual development of phonetic correspondences changed the situation. Rasmus Rask is particularly significant in this transition. Although he too frequently speaks of “letters,” Mounin notes that his work from 1814 already displays a striking attention to systematic phonetic correspondences, including consonantal changes that would later be associated with Grimm's law.
The nineteenth century thus changes the problem. The question is no longer simply how to classify the sounds of one language with great precision. Linguists increasingly need to determine which correspondences recur across languages and what kinds of units can account for them.
This shift matters for the history of the phoneme because the object of linguistic analysis is no longer necessarily something directly available to observation. Historical linguistics works with forms preserved in different languages and attempts to reconstruct relationships among them. The linguist may therefore be dealing with a linguistic unit whose status depends on a pattern of correspondence rather than on direct access to a particular sound.
This development prepares the ground for Saussure. He inherits a linguistic tradition in which sound is increasingly treated as an object of systematic investigation, but he also confronts the more difficult question of what kind of entity a linguist is entitled to posit when the object of analysis is not directly observable.
The transition, then, is not simply from “concrete sound” to “abstract unit.” The problem itself has changed. Linguists are no longer asking only what properties sounds possess. They are also asking what kind of linguistic object can account for systematic relationships among sounds across time and across languages.
That is the problem with which the young Saussure's work on the Indo-European vowel system begins.
3. Saussure Before the Cours: Reconstructing Sounds That Cannot Simply Be Heard
The question of whether the phoneme is concrete or abstract becomes especially interesting when we turn to the young Ferdinand de Saussure. At first sight, his early work seems to belong to a very different linguistic tradition from the structuralism for which he later became famous. The Mémoire sur le système primitif des voyelles dans les langues indo-européennes, published in 1879 when Saussure was only twenty-one, is a work of comparative-historical linguistics concerned with reconstructing the vowel system of Proto-Indo-European. Yet its importance for the present discussion lies in the way it forces us to distinguish a linguistic unit from any single sound that can actually be heard.
The title of the Mémoire already contains a clue: Saussure is interested in a system. His immediate object is the problem of Indo-European vowels, but the investigation cannot be confined to the physical quality of an isolated vowel. The forms preserved in the daughter languages can only be understood by considering the relationships among them and reconstructing the organization that could account for their correspondences.
This distinction is fundamental. A historical linguist cannot listen to Proto-Indo-European. The reconstructed system is not available as an acoustic object. What the linguist has instead are forms preserved in later languages, together with patterns of correspondence and alternation. Saussure's hypothetical elements therefore acquire their status not because they can be directly heard but because they are required by the system being reconstructed.
This does not mean that Saussure had already arrived at the mature structuralist concept of the phoneme. The Mémoire belongs to a historical and comparative framework, and its immediate questions are different from those of the Cours. Nevertheless, something important is already visible: the linguistic reality under investigation cannot simply be identified with the physical substance of a sound. The relevant entity has to be inferred from relationships among observable forms.
John Joseph emphasizes precisely this feature of the Mémoire: its central achievement lies not in isolating an individual sound but in reconstructing a system of vowels and the interrelationships among its elements. The reconstructed element is meaningful because of the explanatory role it plays within that system.
The distinction becomes especially vivid in the account of Saussure's correspondence with his grandfather, Alexandre-Joseph de Pourtalès. Joseph describes Saussure explaining his theory in terms of reconstructed elements whose status could not be reduced to familiar, directly observable vowel sounds. The grandfather's response raises, almost in anecdotal form, the question that later theories of phonology would have to confront: are such reconstructed elements actual sounds, or do they possess a theoretical linguistic value that cannot simply be equated with an observable pronunciation?
The importance of the anecdote is not whether it provides a precise definition of Saussure's theory. It illustrates a more general problem. A linguist may have good reasons to posit an element even when no speaker can provide a directly observable specimen of it. The element is not arbitrary; its status derives from the explanatory structure in which it participates.
This is why it would be misleading to conclude that the young Saussure had already conceived the phoneme as a purely abstract entity. His concern remained historical reconstruction, and his terminology belonged to the nineteenth-century comparative tradition. What we can say more cautiously is that the Mémoire makes it difficult to identify a linguistic unit with a particular physical sound. The linguistic object emerges through a pattern of relations that makes the observed forms intelligible.
This development differs from the approach encountered in the Indian grammatical tradition. The Indian analysis begins with the systematic classification of speech sounds and their recurring properties. Saussure's early comparative work begins with forms preserved across languages and asks what underlying organization is necessary to account for their relationships. The two traditions therefore approach the problem from different directions.
The difference is important. The issue is not that Saussure suddenly abandons material sound in favour of abstraction. The reconstructed system depends entirely upon phonetic and linguistic evidence. What changes is the status of that evidence: observable forms become the basis for inferring a linguistic organization that cannot itself be directly observed.
The Mémoire therefore gives us an important stage in the history of the problem. Before asking what a phoneme is within a synchronic linguistic system, we encounter the more basic question of how a linguistic entity can be established when it is accessible only through a network of evidence and relations.
The answer to that question prepares the way for Saussure's mature theory.
4. Saussure in the Cours: The Phoneme Has No Positive Essence
If the Mémoire shows Saussure working with linguistic entities that must be reconstructed from relations among observable forms, the Cours de linguistique générale turns the problem into a general principle of linguistic theory. The issue is no longer simply how to reconstruct an earlier vowel system. It is what makes a linguistic sound unit what it is within a language.
Saussure's formulation is uncompromising:
“Each language constructs its words out of some fixed number of phonetic units, each one clearly distinct from the others. What characterises those units is not … the specific positive properties of each; but simply the fact that they cannot be mistaken for one another. Speech sounds are first and foremost entities which are contrastive, relative and negative.”
The crucial word here is “negative.” Saussure is not denying that speech sounds have positive physical properties. They can be articulated, acoustically measured, and described in terms of place, manner, voicing, duration, and so on. His claim concerns something different: these properties do not, by themselves, determine the linguistic identity of a unit.
A phoneme, in Saussure's sense, cannot be defined simply by asking what it physically is. We must ask how it differs from the other phonological units of the language.
This is why Saussure's examples of pronunciation are important. A French r may be pronounced in different ways without ceasing to function as the same linguistic unit. What matters is not that every speaker reproduce an identical physical sound, but that the realizations remain sufficiently distinct from the other units with which r contrasts. Conversely, a language may treat a relatively small phonetic difference as significant when that difference serves to distinguish two units.
The consequence is a distinction that will remain central to the rest of our discussion:
A physical difference is not necessarily a linguistic difference.
Two pronunciations can differ materially without constituting different phonological units. Conversely, a phonological contrast may depend upon a very specific phonetic distinction that another language does not recognize as contrastive. The linguistic relevance of a physical difference therefore depends upon the organization of the system.
This is what makes the phoneme relational. It is not an isolated positive entity whose complete identity can be established independently and only afterward inserted into a system. Its linguistic identity is constituted through its relations to the other units with which it contrasts.
Saussure's famous statement that in language there are “only differences without positive terms” becomes clearer when applied to this problem. A phoneme does not possess a linguistic identity independently of the differences through which it is distinguished from other phonemes. The system does not merely record identities that already exist; it provides the relations through which those identities become linguistically relevant.
A useful analogy is the value of a coin. Its value does not derive solely from the metal of which it is made. What matters is the position it occupies within a monetary system. Saussure uses this analogy to insist that linguistic values likewise cannot be identified with their material vehicles. A physical sound is the material medium through which a linguistic distinction is realized, but the linguistic value of that distinction belongs to the system.
This is why it would be misleading to describe Saussure's phoneme simply as “abstract” if abstraction suggests something detached from physical speech. Saussure does not eliminate the material dimension of language. He distinguishes it from the relations that give material differences their linguistic function.
A useful distinction can therefore be made between two questions:
Phonetic question: What properties does this sound have?
Phonological question: What distinctions does this language make among sounds?
The first concerns the material characteristics of speech. The second concerns their linguistic function. Saussure's innovation lies in insisting that the second question cannot be reduced to the first.
This also explains why variation in pronunciation does not necessarily threaten linguistic identity. Speakers do not produce identical acoustic events every time they pronounce what linguists identify as the same phoneme. Yet communication does not collapse, because not every physical difference affects the system of contrasts.
The point is stronger than the simple claim that phonemes are variable. Variability reveals something about the nature of the linguistic object itself. What remains constant is not necessarily a single physical substance but the distinction that the language maintains across different realizations.
The young Saussure's Mémoire can now be seen in a different light. There, the problem was historical: what must be reconstructed in order to make sense of observed vowel correspondences? In the Cours, the problem becomes synchronic and general: what makes one linguistic sound unit distinct from another? The answer is now formulated explicitly in terms of difference, value, and system.
This is where the idea of a positive “essence” becomes problematic. If by essence we mean a property that belongs to the phoneme independently of its relations, Saussure's answer is negative. There is no phonological identity hidden inside a sound that the linguist simply has to uncover. The relevant identity is constituted through the distinctions organized by the language.
Yet this does not mean that physical properties are irrelevant. A language can exploit differences in voicing, aspiration, nasality, length, or place of articulation. What matters is that these differences acquire phonological significance only when they participate in the system.
Saussure therefore leaves us with a precise tension. The phoneme is not reducible to a physical event, but neither can it be understood independently of the material differences through which the linguistic system is realized. Its identity is relational, while its realization remains material.
This tension will become productive when we turn to Jakobson. If linguistic identity is differential, a new question arises: can the differences themselves be analyzed? If a phoneme is defined through opposition, what are the recurring dimensions along which such oppositions are organized?
Jakobson's theory of distinctive features takes up precisely this problem.
5. Jakobson: If the Phoneme Has No Essence, What Are Its Features?
Saussure leaves us with a problem. If a phoneme is not defined by a set of positive properties that belong to it independently of the system, how can we describe it? Saying that /p/ differs from /b/ tells us something important, but it does not yet tell us along what dimension they differ. If phonological identity is differential, does this mean that the physical characteristics of speech have become irrelevant?
Jakobson's theory of distinctive features can be understood as an attempt to answer this question. The move is not a return to the idea that a phoneme has an intrinsic essence. Rather, Jakobson asks whether the oppositions that constitute phonological systems can themselves be analyzed into a limited number of recurrent dimensions.
In Preliminaries to Speech Analysis, written with Gunnar Fant and Morris Halle, the aim is to identify the discrete components of speech and investigate their articulatory, acoustic, and perceptual correlates. This creates a productive position between two apparently opposed approaches. On one side are the material properties of speech: voicing, nasality, aspiration, place and manner of articulation, and so forth. On the other is the structuralist insistence that linguistic identity is constituted through difference. Distinctive features provide a way of relating these two levels.
Consider a simple contrast such as /p/ and /b/. From a phonetic point of view, the two sounds share certain properties while differing in others. Both are bilabial stops; one is voiceless and the other voiced. In a feature-based analysis, this difference can be represented by a feature such as [±voice].
But what exactly has been gained by describing the contrast in this way?
The phoneme has not been reduced to the acoustic event produced by a speaker. Nor has the physical property of voicing been treated as sufficient, by itself, to define the identity of /p/ or /b/. The feature represents a dimension along which the language organizes a phonological opposition.
This distinction is essential. A distinctive feature is not simply a physical property. Voicing exists as a physical phenomenon whether or not a particular language uses it to distinguish phonological units. It becomes distinctive when a linguistic system treats the relevant difference as significant for its organization of contrasts.
This is why Jakobson should not simply be described as returning to a more “concrete” conception of the phoneme. The physical correlate matters, but its linguistic status depends upon the system in which it functions. The feature is neither an arbitrary formal symbol nor a piece of physical substance that exists as a phoneme by itself.
This also helps clarify the relation to the Indian grammatical tradition. Pāṇini and the Indian grammarians classified sounds according to dimensions such as voicing, aspiration, nasality, and place of articulation. The resemblance to feature theory is striking. But the resemblance should not be turned into a historical identity.
In the Indian grammatical tradition, these dimensions form part of a highly systematic classification of Sanskrit sounds. In Jakobson, the aim is to identify a relatively small inventory of recurrent distinctions that can account for phonological oppositions more generally. The two approaches therefore converge in their attention to recurring dimensions of sound, but they organize those dimensions within different theoretical frameworks.
Jakobson's relation to Saussure is equally important. Saussure had argued that speech sounds, considered as linguistic entities, are “contrastive, relative and negative.” A feature-based theory might seem at first to move in the opposite direction. Instead of describing /p/ merely through its difference from other sounds, Jakobson gives us a vocabulary of properties with which that difference can be represented.
Has Saussure's “negative” conception therefore been abandoned?
Not quite. The feature is not a positive essence hidden inside the phoneme. It is a recurrent dimension of difference. [Voice] matters when voicing participates in a relevant opposition; [nasal] matters when nasality functions as a phonological distinction. The feature gives us a way of describing the dimensions along which differences can be organized without making those dimensions independent of the system.
This allows us to formulate the relation between Saussure and Jakobson more precisely. Saussure establishes the principle that phonological identity is differential. Jakobson asks what kinds of differences recur across phonological oppositions and how those differences can be represented systematically.
That is not simply a disagreement between abstraction and concreteness. It is a change in the question.
Instead of asking:
What positive property constitutes this phoneme?
Jakobson asks something closer to:
Through which recurrent dimensions of difference can phonological units be distinguished?
The distinction is decisive. A feature can have articulatory, acoustic, and perceptual correlates without becoming an independent physical essence of the phoneme.
This also explains why feature is preferable to property in this context. A physical property belongs to a sound whether or not a language makes use of it as a contrast. A distinctive feature is a property considered as part of a phonological description: a way of representing a difference that has become relevant within a linguistic system.
The Prague School's development of phonology provides an important context for this position. Phonology was increasingly conceived as the study of functional oppositions rather than simply as an inventory of speech sounds. Jakobson's work develops this structural conception by providing a more explicit vocabulary for representing the dimensions of those oppositions.
The result is not a simple movement from “abstract” to “concrete.” Jakobson's contribution is more interesting than that. He shows how a relational conception of phonological identity can coexist with an analysis in terms of recurrent dimensions that have physical correlates.
The problem that remains is therefore no longer whether the phoneme has an essence. The more useful question is what kind of abstraction a phoneme represents when its identity is relational but the differences through which it is realized have material correlates.
That brings us back to the student's original question.
6. So, Is the Phoneme Concrete or Abstract?
The student who asks whether the phoneme is concrete or abstract is asking a legitimate question. The difficulty is that the two terms conceal several different distinctions.
A particular pronunciation is concrete. It occurs at a particular moment, is produced by a particular speaker, and has particular acoustic and articulatory characteristics. But a phoneme is not identical with any one such realization. No speaker produces exactly the same acoustic event every time a particular phoneme occurs.
In this sense, the phoneme is abstract: it is a linguistic category rather than a single physical token.
But that answer is not sufficient. If “abstract” suggests something detached from physical speech, it does not adequately describe the phoneme either. Saussure shows why: linguistic identity depends upon distinctions that are realized through actual sounds. Jakobson makes the point even more explicit by analyzing phonological oppositions through features that can be related to articulatory and acoustic properties.
We therefore need to distinguish at least three levels:
- The physical realization: an individual occurrence of speech, with concrete acoustic and articulatory characteristics.
- The phonological unit: a linguistic category whose identity is maintained across different realizations.
- The distinctive feature: a dimension of contrast through which phonological differences can be represented and which may have identifiable physical correlates.
The distinction can be put simply:
A realization is concrete; a phonological unit is abstract; the abstraction is grounded in recurrent properties of speech without being reducible to any particular occurrence.
This formulation is useful for the student, but it still leaves an important question. Is there really one object called the phoneme that these three traditions simply describe at different levels?
The history examined so far suggests that we should be more cautious.
Pāṇini's analysis is concerned with the systematic organization of speech sounds and their recurring properties. Saussure's mature theory asks what makes a sound distinction linguistically significant within a system. Jakobson analyzes recurrent dimensions through which such distinctions can be represented. These approaches overlap, but they do not construct exactly the same theoretical object.
This is why the meaning of the term phoneme cannot be separated entirely from the conceptual system in which it appears. A theoretical term derives part of its value from its relations to the concepts with which it is combined. In the limited sense proposed in the Introduction, we can think of each theory as a kind of “mini-langue.”
The analogy becomes clearer if we compare the conceptual surroundings of the term. In the Indian tradition, speech sounds are organized through categories of articulation and sound classification. In Saussurean theory, the relevant vocabulary includes difference, value, relation, and system. In Jakobson's framework, the phoneme is linked to opposition and distinctive features.
The word may remain the same while the theoretical object changes.
This does not mean that the theories are interchangeable, or that every definition of phoneme is equally valid. It means that we should ask what problem a theory is trying to solve and which distinctions it considers fundamental.
The opposition between “concrete” and “abstract” therefore helps us only up to a point. It tells us that a phoneme is not identical with an individual physical event, but it does not tell us enough about what makes the phoneme a linguistic object.
The more revealing question is: Concrete or abstract from which point of view?
From the point of view of phonetic realization, we are dealing with observable physical events. From the point of view of Saussurean linguistic value, the relevant identity is relational. From the point of view of feature theory, the organization remains abstract but can be analyzed through distinctions that have concrete phonetic correlates.
These are not simply three answers to the same question. They are different ways of deciding what aspect of the sound system matters.
And this is where the discussion takes a final turn. If different theoretical approaches bring different aspects of the same phenomenon into focus, perhaps the question is not only what a phoneme is. We must also ask how the phoneme becomes an object of linguistic knowledge in the first place.
7. The Point of View Creates the Object
We can now return to Saussure's methodological principle: “The point of view creates the object.”
At this stage, the statement can be understood more precisely. It does not mean that linguists simply invent their objects. Nor does it mean that physical reality disappears once a theory is adopted. Rather, a phenomenon becomes a scientific object through the questions, distinctions, and relations by which it is approached.
Language contains an enormous number of physical, perceptual, historical, and functional phenomena. A theory selects some of these as relevant, organizes them according to particular relations, and thereby constructs an object of knowledge.
The history of the phoneme makes this process unusually visible.
The Indian grammatical tradition directs attention toward the systematic classification of speech sounds. Its categories make certain articulatory and phonetic distinctions analytically relevant. The object that emerges is therefore one that can be described through recurring properties and organized sound classes.
The young Saussure approaches language through historical reconstruction. His problem is not simply what sounds exist, but what linguistic entities must be posited to explain correspondences among forms preserved in different languages. The object emerges through a network of historical relationships.
The mature Saussure changes the level of analysis again. The central question becomes what makes a sound unit linguistically distinct. The object is now defined through the relations and oppositions within a linguistic system.
Jakobson introduces another level. Once phonological identity is understood as differential, the differences themselves become available for analysis. Distinctive features allow recurrent dimensions of opposition to be represented while remaining connected to the material characteristics of speech.
The important point is not that one of these perspectives is simply correct and the others are mistaken. Each constructs a particular object because each asks a different question.
This is also why the historical relationship among these traditions should not be represented as a simple progression:
Pāṇini discovers the phoneme → Saussure abstracts it → Jakobson completes the theory.
That story would be both historically misleading and theoretically too simple. Pāṇini did not formulate a modern theory of the phoneme, Saussure did not merely abstract a previously existing physical object, and Jakobson did not simply restore the concrete properties that Saussure had excluded.
What we have instead are different constructions of the problem.
Pāṇini's tradition asks how speech sounds can be systematically classified.
Saussure asks what makes linguistic sound units distinct and valuable within a system.
Jakobson asks how the dimensions of those distinctions can themselves be analyzed and represented.
The differences among these approaches do not make the phoneme unreal. They show that the linguistic object is not exhausted by any single level of description.
This is where the “mini-langue” analogy becomes useful. A theoretical term acquires part of its meaning from the relations that surround it. Just as a linguistic sign has no value in isolation, a theoretical term cannot be fully understood in isolation from the conceptual system in which it functions.
The analogy should not be pushed too far. A linguistic theory is not literally a langue, and theoretical concepts do not function exactly like linguistic signs. But the comparison captures something important about the problem we have been following: the identity of a theoretical object depends partly on the network of distinctions through which it is constituted.
The result is not relativism. It does not mean that any definition of phoneme is equally valid, or that linguistic theories simply manufacture arbitrary realities. Different theories are constrained by linguistic evidence, empirical observation, explanatory requirements, and the phenomena they seek to account for.
What changes is the object that becomes theoretically salient.
The point of view does not create speech. It does not create the acoustic event, the articulatory gesture, or the historical correspondence. But it creates the particular object of linguistic analysis that emerges when those phenomena are organized under a specific theoretical question.
This gives the student's original question a different significance. Asking whether the phoneme is concrete or abstract assumes that there is a single object whose ontological status can be settled independently of the theory through which it is approached. The history we have examined suggests otherwise.
The better question is not simply:
Is the phoneme concrete or abstract?
It is:
What kind of object becomes a phoneme when we look at speech from a particular point of view?
Conclusion: Three Phonemes?
We began with a simple question: Is the phoneme concrete or abstract? The history we have followed suggests that the question cannot be answered independently of the theoretical object being constructed.
Pāṇini's tradition shows how systematically speech sounds can be classified through recurring articulatory and phonetic distinctions. Saussure shifts attention toward the relations that make sound units linguistically significant. Jakobson asks how those relations can themselves be analyzed through recurrent distinctive features.
These are not simply three competing definitions of one already constituted object. They are three ways of approaching what counts as relevant in a sound system.
Perhaps, then, there are not three answers to the question “What is the phoneme?” but three different ways of constructing the question.
The student's question remains useful because it exposes the tension between physical realization and linguistic identity. But it does not exhaust the problem. The more fundamental question is:
Concrete or abstract from which point of view?
Saussure's methodological maxim provides the final clue: the point of view creates the object. The phoneme is not simply waiting to be discovered as either a physical thing or an abstract entity. What counts as the phoneme depends on the theoretical relations through which a linguist approaches the sound system.
The history of the phoneme is therefore not simply a movement from concrete sound to abstract structure. It is a history of changing perspectives on what counts as linguistically relevant in language.
References
Cardona, G. (1997). Pāṇini: His work and its traditions (2nd rev. & enl. ed., Vol. 1). Motilal Banarsidass Publishers.
Jakobson, R. (1962). Selected writings: Vol. 1. Phonological studies. Mouton.
Jakobson, R., Fant, C. G. M., & Halle, M. (1952). Preliminaries to speech analysis: The distinctive features and their correlates (Technical Report No. 13). MIT Acoustics Laboratory.
Jakobson, R., & Halle, M. (1956). Fundamentals of language. Mouton.
Joseph, J. E. (2012). Saussure. Oxford University Press.
Joseph, J. E. (2022). Saussure and structural phonology. In B. E. Dresher & H. van der Hulst (Eds.), The Oxford history of phonology (pp. 203–220). Oxford University Press. https://doi.org/10.1093/oso/9780198796800.003.0010
Mounin, G. (1985). Histoire de la linguistique: Des origines au XXe siècle (4th ed., rev.). Presses Universitaires de France.
Saussure, F. de. (1879). Mémoire sur le système primitif des voyelles dans les langues indo-européennes. B. G. Teubner.
Saussure, F. de. (2011). Course in general linguistics (W. Baskin, Trans.; P. Meisel & H. Saussy, Eds.). Columbia University Press. (Original work published 1916)
Trubetzkoy, N. S. (1969). Principles of phonology (C. A. M. Baltaxe, Trans.). University of California Press. (Original work published 1939)
Comments
Post a Comment