Analysis of a Sentence

It is a useful exercise, for those interested in language, to list the various ways in which a sentence can be analyzed, in order to gain a full appreciation of how multi-faceted a sentence is.[1] Strangely, I have never seen this done. As an example, I will choose the sentence “Miami is hot”, but any sentence will do, since the same classification scheme applies to any sentence in any language. These are to be universals of language. (1) Ethnography: what language does the sentence belong to, who speaks it, and where? The same string could occur in more than one language, and a given sentence can contain words from more than one language. (2) Lexical: what words are contained in the sentence and in what order? Are there any ambiguous words or nonsense words or misspelled words or obsolete words? (3) Phonetics: what phonemes does the sentence (or its utterance) contain? Here we look at the acoustic and articulatory properties of the sentence. These can vary from speaker to speaker and dialect to dialect. Volume and location of utterance are not relevant. (4) Syntax: what is the syntactic or grammatical structure of the sentence? Now things become more theoretical and difficult. As a first approximation, we can say that our sample sentence contains a noun, a verb, and an adjective; and it is a simple subject-predicate sentence (no sentence connectives). It is in the indicative mood. These properties do not vary from speaker to speaker; they are intrinsic to the sentence’s syntax. (5) Semantics: what does the sentence mean? How does its meaning depend on the meaning of the words that compose it? It means that Miami is hot, and it is composed of a singular term denoting the city of Miami and a predicate ascribing the property of being hot to that city. In the case of other sentences, we will identify truth-functional connectives, intensional operators, quantifiers, indexicals, and other devices. We might also speak of logical form, truth conditions, functions from objects to truth-values, propositions, possible worlds, predicate modifiers, etc. (6) Pragmatics (or linguistic use): how does the sentence relate to speakers who use the language? In what circumstances is the sentence used? What are the criteria of assertion for the sentence? What are the likely consequences of asserting it? What language game does it belong to? (7) Conceptual analysis: how are the concepts expressed by the sentence to be analyzed? Here we might claim that the name expresses a definite description picking out the city of Miami, and that the predicate ascribes a dispositional property (apt to feel hot to normal people in normal conditions). We are elucidating the concepts that the sentence expresses, spelling out their underlying content. (8) Ontology (or metaphysics): what is the ultimate nature of the entities that form the subject matter of the sentence? Should we be realists or anti-realists about them? Are they mental or physical (idealism versus materialism)? Are they substances or events? Are they Many or One? Is God essential to their existence? (9) Phenomenology: how does the sentence manifest itself in consciousness? What is the nature of the intentional acts that constitute grasping the sentence consciously? What are the relevant noemata? Is the consciousness of death (Heidegger) part of apprehending the meaning of the sentence? (10) Neurological: what are the brain processes that underlie understanding the sentence? How is its structure represented in the brain? In what ways can brain damage impair using and understanding the sentence? (11) Psycholinguistic: what is the cognitive psychology of the sentence (any sentence)? How is it produced by the speaker and processed by the hearer? What is the role of attention in speech comprehension? What causes performance errors? How is the sentence acquired by the child, and when? (12) Psychoanalytic (if any): how does the sentence bear on the neuroses, dreams, and complexes of the speaker or hearer? How does it interact with the unconscious mind, the id, the superego, repression, sexual development? (13) Literary: what are the literary aspects of the sentence (here we might switch to “To be or not to be”)? Does it contain rhymes, assonance, onomatopoeia, poetic depth? Or is it banal, cliched, derivative? (14) Etymology: what is the history of the words composing the sentence? From what language do they derive? Is their current meaning close to their original meaning? (15) Cachinnation: is the sentence funny? Is the word combination amusing? Is it in bad taste? Would you say it in front of your mother? Is it the punchline of a joke? These are all legitimate questions we can ask about a sentence (there may be others). They all involve analyzing the sentence (OED “analyze”: “examine methodically and in detail for the purposes of explanation and interpretation)”. Some are more controversial than others. The philosophically relevant kinds of analysis occur naturally on the list along with the other kinds. Sentence analysis brings in linguists, psychologists, philosophers, physiologists, psychotherapists, literary scholars, historians of language, and humorists. The sentence is a many-sided creature, not the exclusive property of any one field. No single field can claim to comprise the whole of Sentence Studies. For the philosopher, it is salutary to observe that many fields seek to analyze sentences, each for its own purposes; philosophers are just doing more of the same, but with respect to different dimensions of the phenomenon. Sentences are subject to conceptual analysis, ontological analysis, and phenomenological analysis, as much as phonetic, syntactic, and physiological analysis. Perhaps it is true to say that language is susceptible to more types of analysis than any other phenomenon of nature. Or to put it differently, it is the most interesting and complex thing in the world.[2]

[1] I will leave aside the question of what a sentence is. It is certainly not sound waves in the air or marks on paper; such things could arise in any number of ways and have nothing to do with language. Sounds and marks are vehicles of sentences not sentences themselves. Are they something psychological or neurological or computational or abstract? It is hard to say. People tend to assume that sentences are clear and meanings unclear, but really sentences are unclear too. What makes a sentence what it is—a syntactic object capable of bearing meaning? Here is where the idea of a language of thought becomes attractive, because such items would have syntax built into them ab initio. The ontology of sentences is obscure (like that of propositions).

[2] I write this partly because there is currently a fashion for decrying conceptual analysis in philosophy. But really analysis of sentences is customary across many fields, and is not to be dispensed with. See also my “Analysis of Analysis”. It is certainly not to be equated with something pejoratively labeled “ordinary language philosophy”.
Analysis is an essential part of science.

Analytic Philosophy as Phenomenology

Phenomenology lies in a long tradition stemming from Descartes and including Locke, Berkeley, Hume, Kant, Mill, Brentano, and Wundt. It has a number of distinguishing features, as developed by Husserl: it is based on “intuitions”; it is concerned with phenomena not hidden realities (e.g., Kantian noumena); it aims to discover essences; it combines elements of empiricism with elements of rationalism; it aspires to be a science; it makes heavy use of the notion of intentionality; it is introspective; it employs something called the “phenomenological reduction”, which seeks to suspend the “natural standpoint”, in a mental act called the “epoche”. The first and last items in this list require special mention. In phenomenology, the concept of intuition is used in a broad and specialized sense: it is taken to include the full range of basic acts of consciousness, from perception to imagination to abstract thought, so that a physicist is said to use intuitions when making perceptual observations. What is necessary is that an intuition involves confrontation with the object of interest not merely indication in absentia; this can be perceptual or imaginative. It is to be contrasted with inference or argument or abstract speculation. It is a source of data. The idea of the phenomenological reduction is that we are to step out of the usual attitude of science and common sense, which involves commitment to external reality, and contemplate only the contents of consciousness as such; we are to be concerned with essence not existence. It is a matter of indifference whether the objects of consciousness exist (it is not denied that they do). In practice this means that we are concerned with the mind as opposed to objective reality. Thus, we elicit essences by suspending the natural standpoint and employing intuition as our source of evidence. We attend to what is given from the first-person perspective in order accurately to describe the phenomena. We don’t rely on theology or traditional authorities or science or mysticism or a priori deductions. We use an empirical method in order to arrive at necessary truths (this is called “eidetic abstraction” by Husserl). We are operating at the same level as Descartes in his search for clear and distinct ideas, and also with the British empiricists in their preoccupation with ideas, images, perceptions, impressions, sense data, and the like. Our gaze is directed inward, but we take in more things than they recognized (numbers, for example). Now my question is this: Isn’t this very similar to what the analytic philosophers recommended? That is, didn’t they take a similar stance towards the investigation of concepts? They proposed to examine concepts by detaching them from the world and intuiting them (in the wide sense) so as to reveal necessary truths. They wanted to analyze these concepts just as Husserl proposed to analyze his “noema”. They were not concerned to discover empirical and contingent truths about the extensions of these concepts but to articulate the essence of the concept (as it might be, the concept of knowledge). In some cases, this could be accomplished by simply looking at the concept (e.g., reporting that the concept of red is a color concept), while in other cases a more complicated procedure would need to be adopted—the imaginative construction of thought experiments and provision of necessary and sufficient conditions (compare Husserl’s method of eidetic variation). We thus gain insight into the “logical structure” of the concept in question. We don’t need to look outside of the mind in order to do this (or if we do, it is only to look at language); we don’t mimic the natural sciences by investigating the extra-mental natural world. We have “bracketed” that mode of enquiry so as to concentrate on essence. Since concepts are “immanent” in the mind, we are not then subject to skeptical doubts, so that we have a firm foundation for our enquiries. We are, in effect, conducting a phenomenology of concepts, in Husserl’s sense. Analytic philosophy is therefore a species of phenomenology, and the better off for it. Concepts are phenomena that can be studied by means of intuition, supplemented by acts of imagination, and yielding knowledge of necessary truths. Analytic philosophy is not opposed to phenomenology but a type of phenomenology: first-person, introspective, intuitive, non-scientific, apodictic, and essence-seeking. The differences start to seem merely terminological and stylistic. This impression is confirmed by asking a further question: What kind of meta-philosophy stands in contrast to that of analytic philosophy and phenomenology? Here we cannot do better than to tabulate Sartre’s deviations from Husserlian orthodoxy. These concern his conception of consciousness itself: he regards consciousness (the for-itself) as a nothingness whose only qualities are conferred by extraneous being (the in-itself). From this it follows that there can be no detachment from existence, no suspension of the natural standpoint, no epoche: the for-itself is constituted by the in-itself. There is no transcendental ego in consciousness and no “stuff” of consciousness (Husserl’s “hyle”): consciousness is pure intentionality and hence dependent on the world of actual existence. It is, for Sartre, absolute freedom, devoid of essence, a kind of psychic vacuum. Thus, Sartre describes consciousness as embedded in the existing world—in the world of physical objects, the human body, time, and other consciousnesses.[1] The division between them is artificial; it is the consciousness-world nexus that is ontologically basic in the description of human reality—what we might call “situated consciousness”. What does this remind you of? Externalism, anti-individualism, the extended mind, wide content, Twin Earth, direct reference, singular propositions, meaning outside the head, belief de re, non-supervenience of mind on brain—all that jazz. This is all very Sartre-esque and not at all Husserlian: if you suspend the world, you do away with the mind. Jean-Paul would have loved the Twin Earth story! Not for him the neo-Fregean insistence on internal modes of presentation, definite descriptions in the language of thought, narrow content, supervenient qualia—all that internalist claptrap! Husserl is to Frege what Sartre is to Mill. The analytic externalists are “existentialists” in that they build objective existence into the mind; they reject Husserl’s attempt to insulate the mind from the outside world. The mind is embedded and extended, not isolated and detached. And there is a further point of analogy: Quine’s view of the mind as indeterminate is strikingly similar to Sartre’s vacuum picture of the mind. For both men, the mind is essentially an empty vessel, consisting of nothing but interactions with the external world. It brings nothing to the table; it merely reflects what is already there. Sartre would have been tickled by Quine’s rabbit ruminations: for there is nothing in the mind to provide any content beyond what can be gleaned from stimulus meaning. The in-itself might well not provide the distinctions necessary to justify our customary discriminations, in which case consciousness cannot furnish such discriminations. This kind of picture of the relationship between mind and world upends centuries of thought about what constitutes the mind—in particular, the views of Husserl’s forerunners (Descartes, Locke, et al). So, the departure represented by Sartre and the psychological externalists is really a major divide in the historical tradition—more significant than the one gestured at by the usual hackneyed “analytic versus Continental” dichotomy. In fact, the former division cuts across the latter division. All can still count as phenomenologists, but they differ dramatically in how they conceive their object of study (the mind, experience, consciousness). There are existential phenomenologists (Sartre) and non-existential phenomenologists (Husserl). And the same distinction applies to conceptual analysts, according to whether they regard concepts as existence-involving (Putnam) or not existence-involving (Frege).[2] It turns out, then, that analytic philosophy and phenomenology are not really opposed, despite some superficial stylistic differences. Both derive from the same historical sources (Descartes, Hume, et al), so this is not surprising. It is the focus on the knowing subject that leads to both.[3]

[1] Sartre disavowed the label “existentialist” (it was Gabriel Marcel that coined it)—as did Camus. What would be a better label? We might try “existence-ist”, or “libertarian negationist”, or simply “externalist”: but these are not too catchy or descriptive. Still, we do well to remember that “existentialist” does not do justice to the guiding principle of Sartre’s philosophy, namely the essential nothingness of consciousness—he is a “nothing-ist” more than an “existential-ist”. I rather like “phenomenological negativist” (contrast “logical positivist”), but it’s a mouthful. (Apparently, Husserl used to like to say “We are the real positivists!”, and with some justification—they were a pretty negative lot.)

[2] Where would Wittgenstein fall? He can be aptly characterized as a phenomenologist of sorts (“Look and see!”), but what about along the internalist-externalist axis? Well, he isn’t a referential externalist (nothing twin-earthy about PI), but he does invoke “forms of life” and the community, so he is not an it’s-all-in-the-head type of philosopher. He also shares with Sartre an attachment to the idea of life as decision (see section 186).

[3] A view which really does stand opposed to both phenomenology and analytic philosophy is that philosophy is not different from regular science, since both claim that philosophy needs a special method. What should we call such a view? To call it “naturalistic philosophy” presupposes that the other two approaches are not “naturalistic”, but their proponents would be within their rights to claim that they are perfectly naturalistic (they aren’t “super-naturalistic”). Nor would it be apt to call the view “scientific philosophy”, since the opposition would protest that they too deserve that honorific adjective. How about “scientistic philosophy”? But that connotes the fault of excessive belief in scientific methods: true, no doubt, but not diplomatic. I think “scientistical” would do, or “scientifical”. Thus, scientifical philosophy would differ from phenomenological philosophy (which includes analytic philosophy)—which in turn differs from religious or mystical philosophy, or psychedelic or alcoholic or schizophrenic…

Intentionality and the Ego

In The Transcendence of the Ego (1936) Sartre criticizes Husserl’s conception of consciousness. I intend to add to this critique. Husserl supposes that in addition to the usual objects of consciousness there exists a further object christened the “transcendental ego”: it is the reference of “I”, the source of the unity of consciousness, and a constituting agent with respect to the intentional objects of consciousness (e.g., the objects of perception). This implies that consciousness inherently possesses a dyadic structure, a bipolarity, a subject-object axis: I am aware of this. It has a kind of double reference—to the ego and to the objects the ego contemplates. Sartre contends that this is false: there is no ego embedded in consciousness per se, only the objects that consciousness apprehends outside itself. The ego enters, according to Sartre, only when consciousness acts reflectively, taking itself as object; it doesn’t enter at the basic level of intentional directedness—as when you see or think about things in general. Consciousness necessarily posits (Sartre’s term) the objects of its essential intentionality, but it only sporadically posits an ego, as when it turns back on itself. The ego is not part of the constitutive structure of consciousness. It might well be true that there has to be an ego (self, subject) in order for consciousness to do what it does, but this is not part of its essence quaphenomenological reality. There no doubt has to be a body and brain in order that consciousness should achieve its essential intentional directedness, but this is not part of its phenomenology. Consciousness is not as of an ego that transcends it (i.e., non-reflective consciousness). Now, I think Sartre is quite right in this contention, and that his case can be strengthened. Consider this analogy: if I say “New York is noisy” I refer to New York and its acoustic properties, but I don’t refer to myself. I am there in the background, a necessary condition of this particular act of speech, but there is no “I” in my sentence. I can go on to say “I just said that New York is noisy”, thus engaging in a self-reflective act, but I don’t typically do that. It is the same with consciousness: I can think about the world without thinking about myself in having that very thought. Speech is not invariably self-positing, and neither is thought (or perception, emotion, etc.). Moreover, if (first-level) consciousness did posit an ego alongside its other posits, it would be possible for it to exist without those other posits; but that is clearly impossible, because it would be empty of what gives it reality (by the principle of intentionality). It would be the mere apprehension of the solitary ego devoid of reference to anything outside of consciousness. But consciousness is necessarily about something distinct from itself, of something separate and alien. The idea of a consciousness containing only reference to the pure ego is not intelligible (Husserl would agree, given his allegiance to Brentano). The transcendent ego is not enough transcendence (“going beyond”) to sustain an act of consciousness. Secondly, if the ego were contained in consciousness as an intentional object (and how else is it to be contained?), there would have to be another ego to be the subject of that apprehension, unless we thought that it could be aware of itself. But intentionality is inherently irreflexiveby definition: this would be an inhabitant of consciousness being directed at itself, like a perception being a perception of itself. Not surprisingly, Husserl didn’t postulate that his transcendent ego occurred in consciousness by way of the intentional relation, but rather in some other way—as a “constituent”. But this violates the whole doctrine of intentionality, and is anyway obscure. The fact is that the way consciousness gives access to the ego is (or would be) quite different from the way it gives access to other objects, but this difference is never explained. Thirdly, if the ego has a nature, this nature would have to constitute part of the nature of consciousness itself; but consciousness has no nature apart from what its objects confer (the doctrine of intentionality again). If the ego is identical to the psychological subject (the “empirical ego”), then consciousness will be invaded by desires, personality traits, and so forth; but nothing like that is found within the precincts of consciousness as such (it is an “emptiness”, as Sartre says). The same goes for the ego considered as a body, a brain, an extensionless point, or an immaterial soul: none of this shows up in consciousness (unless it is explicitly about such things). But the ego can’t have no nature—it can’t be a nothing. The usual objects of consciousness do color the conscious act, constituting it as the act that it is, and that fits with the deliverances of phenomenology; but nothing analogous happens with the alleged transcendental ego—it makes no phenomenological difference. Fourthly, what are we to say of animal consciousness—does it too harbor a transcendental ego? That seems far too intellectualist: isn’t the animal simply aware of objects outside its consciousness, not of any supposed ego lurking within it? Why would its consciousness be directed at something on the inside as well as the outside? What would be the biological point of that? It seems superfluous, a kind of pointless luxury. Fifthly, what guarantees that the putative reference to an ego actually picks out anything real? Suppose we agree that such a directedness occurs: it doesn’t follow that it points to anything that really exists. So why should we believe in such a thing? It might an illusion; and wouldn’t an illusion do just as well as the real thing, phenomenologically speaking? As Sartre argues, we do better to regard consciousness in its basic form as unipolar—pointed solely at its objects, not as containing in addition some kind of reference to a self. Visual perception, say, represents objects in the environment without any accompaniment by a representation of some kind of ego (whose nature remains obscure). There is no positing of a self, analogous to the positing of physical objects in the environment. No doubt there has to be some kind of entity underlying the acts of consciousness—a person, an organism, a brain—but this entity doesn’t show up in the phenomenology of the act. Why should it? The being of consciousness qua consciousness is intentionality not constitution by a supposed inner ego. That idea is really the abandonment of the original insight (derived from Brentano) of the phenomenological movement. The phenomenological field should be kept ego-free.[1]

[1] Of course, thoughts about the self, one’s own or other people’s, contain reference to such things; but this is not a feature of all thoughts. Most of consciousness is quite innocent of self-directedness. Thus, there is no ego-as- subject in consciousness but only ego-as-object (in the special case of reflexive consciousness). There is no mention of the self in a typical conscious act; there is no “I think” accompanying every exercise of consciousness (pace Kant). Consciousnesses thus don’t differ with respect to the individual possessing them; they are “impersonal”, as Sartre puts it.