First published by G. Debrock and M. Hulswit (ed.), Living Doubt, 109-120. 1994, Kluwer Academic Publishers. Printed in the Netherlands.

1 Comte and Peirce

In previous studies I have dealt quite a lot with the philosophy of Auguste Comte, and what strikes me is how much better specialists of Comte would have understood Comte's view, had they known Peirce's semeiotic! In previous publications, I have repeatedly shown that Comte's work displays strong major and fundamental semiotic dispositions (Kremer 1982; 1983; 1988; and particularly Kremer 1983b:19-26). It is obvious to me that some of Comte's theories correspond with those of Peirce, especially in the fleld of epistemology, which Comte approached through a theory of language, history and the philosophy of science. If Peirce has criticized Comte, he might also have perceived some similarity with his own philosophy.

I have in mind particularly what Comte called his "positive logic", which as a set of three semiotic Iogics: the first called the logic of feelings, the second, the logic of images, and the third, the logic of signs. The tripartition of "feelings," "images" and "signs" contains semiotic elements recognizable in Peirce's semeiotic. Indeed, Comte's positive logic may be compared roughly with Peirce's original distinction between Firstness, Secondness, and Thirdness.

From an anthropological point of view, Comte found that fetishism, polytheism and monotheism were at the origin of these three logics, whose development was closely linked to the system of society. He thought that language is composed of connection-signs between two elements, thereby forming a sign system. Thus, man and the world constitute an interdependent system which is the effect of language. Some features of Comte's semiotics correspond with the Saussurian dichotomy, while others write the Peircean trichotomy. Comte proposes the dual distinction sensation/movement and its homologous "inner images"/"outer images;" but, like Peirce, he assumes a third connection-sign between the two elements. Thus, Comte reconstitutes the zoological roots of the sign based on the two sides of the reflex arc combined with a complementary third side constituting the "intermediary vitality".

Beyond a fundamental zooesemiotics, there is also an anthropo-semiotics that determines on semiotics of Comte. Indeed, his epistemology hinges upon a generalized theory of language. It will be shown that the same is true for Peirce. There are certainly important differences between Comte and Peirce. Comte was haunted by the idea of constructing a system capable of helping French politics in the first half of the 19th century, whilst Peirce did not attempt such a system. Comte considered the fulfillment of humanity as the end and symbol of his Positive Philosophy, while Peirce saw humanity as a non-sufficient end. For Peirce humanity belongs to Secondness and reason to Thirdness. Moreover, as neither humanity nor reason was the final end, Peirce implied there was, in the words of Herbert Schneider (Schneider 1952) a kind of Fourthness that was the summum bonum (CP 5.433) [1].

2 The Originality of Peirce 's Epistemology

Peirce opened a paved way in which epistemology is linked to a theory of meaning and a theory of signs. Indeed, the general concern of epistemology is the question of what knowledge is and how it is possible. Epistemology seeks to answer the question: How can humans know about the natural world? Peirce gave his answer: only through the way of signs considered as tools and objects. Signs are not only linguistic, but also logical and pragmatic. Peirce is right to assert that we cannot think without signs (CP 5.251) and that every thought is itself a sign (CP 5.253), so that we are concerned with the concept of thought-sign (CP 5.283) and the idea that even man himself is a sign (CP 5.310).

Thus all signs, whether linguistic, logical or pragmatic, belong to a generalized theory of language, which Peirce summarized very early as follows: "Being, quality, relation, and other universals are not known except as characters of words or other signs, attributed by a figure of speech to things." (CP 5.343)

Since it is accepted today that sciences take symbolical facts as objects to be described and constructed,1 the task of epistemology has been re-defined: one must become aware of the ways in which signs are used as tools and objects in our knowledge, both on the level of ordinary language and of scientific discourse.

In matters of epistemology, Peirce gave us a theory of language, within which he lays bare the origins and the nature of knowledge, the presuppositions of knowledge, and the conditions of veracity of knowledge. But besides epistemology, Peirce developed many other fields of interest, one of which is that of speculative grammar (Réthoré 1988). Indeed, epistemology presupposes the speculative grammar, for the latter studies the signs themselves while epistemology studies the signs in their relation to the world.

The epistemologies of Comte and Peirce alike presuppose a generalized theory of language, which implies a reciprocity between reality and signification. The epistemology of Peirce is governed by an operational principle that acts as a primary, universal postulate which leads every activity of the human mind. In fact, the mind acts as mediator between the order of nature and the order of a community (which may be either a particular community or the community at large); therefore its activity depends upon both nature and society.

In respect of society, Peirce had recognized as early as 1868 - in the concluding portion of his article "Grounds of Validity of the Laws of Logic: Further consequences of Four Incapacities" - the fact that "the social principle is rooted intrinsically in logic" (CP 5.354). And as for nature, he had given in "How to make our Ideas clear" (1878) a kind of natural phenomenology of thought, speaking of doubt as a cause of irritation arising from indecisiveness in matters of action. The function of thought was to produce belief, after it had been excited by doubt. The effect of thought, which is belief, involves a habit which in turn is a "rule of action" so that belief becomes a "habit of action." Peirce's idea that "thought is essentially an action" (CP 5.388-410) is a view which he shares with Hobbes (Hobbes 1966a:5.253;1966b). Now, the action of thought begins with an activity, and this activity is a generalization of the effort (CP 5.442) of any kind of language which may be constituted by images, for percepts are images or kinetic images (CP 5.115).

Doubt and Belief, or question and decision, constitute thought which manifests its reality to us as arising from a stream of images passing rapidly through consciousness. In this regard, Peirce observes that thought "runs" like a melody through the succession of our sensations (CP 5.388-410), a theme which later would be extensively worked out by both William James and Henri Bergson.

Moreover, as thinking beings, men apprehend the world flot through a personal, intuitive cogito (Peirce is strongly anti-Cartesian), but by virtue of the social status of logic. It is because the ideal perfection of knowledge which in the end constitutes reality belongs to a community, that any man within that community has access to the world only in as much as he identifies his own interests with those of the community (CP 5.356).

Furthermore, the human mind may be said to be founded "in the world." I borrow this Heideggerian expression, although it is clear that, within the context of Peirce's work, it should be interpreted in an altogether different manner. In "The Order of Nature" (1878) Peirce wrote "... that the mind of man is strongly adapted to the comprehension of the world ..." (CP 6.417). Peirce establishes this fact which he considers as incontestable on the basis of some sort of holistic principle which probably constitutes Peirce's most basic certitude: that there is a total harmony between man and world. Here is another similarity between this principle and Comte's positivist certitude.

Indeed this very same belief is related to the question put forward by Einstein who wondered why it should be possible for the human mind to understand the universe. But, much though he could not understand why the universe is intelligible, he nevertheless was convinced that it was intelligible. Einstein did formulate his answer to that enigma in terms of the notion of unity, when in a letter to a friend (January 1938) he wrote: "The logically simple does not, of course, have to be physically true; but the physically true is logically simple, that is, it has unity at the foundation" (Holton 1988:259).

Peirce too refers to unity when he suggests that the understanding of the universe may be theoretically demonstrated by virtue of the theorem that ..... there is a character peculiar to every possible group of objects" (CP 6.414). Practically however, the existence of this character may be shown to obtain through the action of thought, i.e., through "mental action," but also through the use of language, or, more precisely, through the use of any sign system and the kind of meaning which such system may provide. Both signs and representations belong to the nature of the real; both share the status of a positive fact, and therefore the forcefulness and hardness of Secondness. But the reality of the real is made possible by thought, i.e., by Thirdness. Thus, there is no "thing in itself" in the epistemology of Peirce, for all reality is basically founded upon interpretants and interpreters. The famous pragmatic maxim provides the rule that governs an epistemology which, instead of presupposing innate ideas, is based on the assumption that there is something like social-historical and experimental knowledge.

"Consider what effects, that might conceivably have practical bearings we conceive the object of our conception to have. Then, our conception of these effects is the whole of our conception of the object." (CP 5.402)

The pragmatic turn of epistemology, expressed in 1878, is profoundly positive, and perfectly coherent with a holistic explanation that is valid for all phenomena. This coherence seems to me so well defined by Robert Almeder in relation to the meaning of sentences: "The meaning of the sentence is issued in terms of the conditions for the whole theory in which the sentence is embedded" (Almeder 1983:346,n16). Indeed, for Peirce, meaning is at once referential, pragmatic and structural: it is a whole, not unlike the entire system "man-world".

3 The Legality of Scientific Communication

The main criticism levelled against Peirce's epistemology challenges the principle that the human mind is adapted to nature, and therefore rejects what has been called "Peirce's Ontological Postulate," viz. "that the structure of logic is the mirror of the structure of reality" (Freeman 1983:70).

This raises the question how Peirce could possibly justify his assertion of the structural homology between logic and reality, according to which the laws of the universe and the laws of the mind are positively the same (a position shared by Comte). The answer may be that man and the world are mediated by a situation-bound language which is the expression of human abilities and the range thereof. Indeed, Peirce teaches us that the semeiotic unity is represented by the proposition. Thus he stated in his "New List of Categories" (1867) that "The Unity to which the understanding reduces impressions is the unity of a proposition" (CP 1.548).

Following Kant and anticipating Einstein, Peirce was confronted with the problem of having to reduce "the manifold of sensuous impressions to unity" (CP 1.545). And he thought that this unity was given by the proposition which is "the connection of the predicate with the subject" (CP 1.548), as thought through the conception of being expressed in the copula. Abstraction or precision constitutes the conjunction of substance and being, and is first rendered possible through the proposition expressing the substance and its quality in the reference to a ground. The conception of quality heralds the passage from being to substance. But in order to know a quality we need a relation within the reference to a correlate. Finally, the reference to a correlate "conjoins to the substance the conception of a reference to an interpretant" (CP 1.553). Therefore, in addition to the related thing with ground and correlate, we need, in order to compare, a mediating representation: the latter has to represent the relate as being a representation of the same correlate represented by the mediating representation itself (CP 1.553). And Peirce explains why this mediating representation must be called an interpretant: "... because it fulfills the office of an interpreter, who says that a foreigner says the same thing which he himself says" (CP 1.553).

Thus, between being and substance there are these three accidents, quality, relation and representation, which from the very beginning are called resp. First, Second and Third, and which would later be developed into the conceptions of Firstness, Secondness and Thirdness. These five conceptions of being, substance, quality, relation and representation constitute the new list of categories which Peirce established without making use of the terms of Aristotle or the judgments of Kant. Instead, he bases them upon the forms of inference according to whether a related thing is referred to a ground, a correlate or an interpretant. Thus, for Peirce, inference is the essential function of our knowing mind (CP 2.444). Peirce saw in these fundamental concepts the three elementary forms of predication or signification: qualities, relations and representations (CP 1.561). The quale (in 1905 he replaced "quality" by "non-relative characters" (CP 1.565) refers to a ground; the relation which, having an object, the related, refers to both the ground and the correlate; and the representamen not only represents the object, but also refers to the ground, the correlate and the interpretant.

Peirce, who was quite proud of having found the Third, claimed that this allowed him to go further than De Morgan, the author of the essay "On the Logic of Relations" (1866). From De Morgan, Peirce had inferred that indecomposable predicates were of three classes: predicates applying to a single subject, predicates having two subjects (i.e., the subject nominative and the object accusative), and those having three such subjects or correlates. But De Morgan had failed to perceive the last of these classes, which according to Peirce expresses "some relation of an intellectual nature" (CP 1.562). All in all, the first Logic of Relations stated in 1867 formed the basis of Peirce's epistemology.

But Peirce went on developing his logical investigations in the field of relations, and as a result he changed his perspective on the nature of the proposition several times. In 1870 in his "Description of a Notation for the Logic of Relatives," he redefined the proposition as a transitive relation of inclusion. In his many writings during the year 1890, however, he maintained his first definition: a proposition consists of two parts, the predicate and the subject (MS 280). In "The Basis of Pragmaticism" of 1906, he observes that truth belongs to the proposition (CP 5.553). He considers the subject to be an indexical symbol, and the predicate an iconic symbol, while these are related to each other by a copula (CP 3.621) which is said to express a relation between some general terms and the universe.

In respect of the link between subject, copula and predicate, this logical trichotomy is remarkably confirmed in Strawson's book Subject and Predicate in Logic and Grammar (Strawson 1974; see also Paliaro 1950), where a similar idea of the tripartition of functions in a propositional combination is proposed. For Strawson, a sentence contains an expression specifying the particular, an expression specifying a general concept, and a third element which is described as "some feature of the mode of combination of the two aforementioned expressions" (Strawson 1974:21). In Strawson's view, it is this combination that "yields truth, if the particular exemplifies the concept - or, if the concept appliesto the particular" (Strawson 1974:21). The first two functions are the same in both logic and grammar, but in respect of the third function, Strawson states that the 'grappling machinery' is located on the verb-phrase, more surely than on the noun-phrase. Moreover, Strawson shares Peirce's view that the notion of sign is at the same time "expression," "meaning," "representation" and "communication." For Peirce the verb is primarily a pure icon, and only then a complete relative and a model of the general functioning of any assertion, while the copula has a function but no object (CP 2.343).

Indeed, apart from stating the tripartition of the proposition, Peirce also shows there is a link between this tripartition and the notion of communication. Any communication presupposes an utterer and an interpreter. Yet the utterance is still dependent upon a particular kind of assurance which is also subject to the principle of trichotomy: "As to the Nature of the Assurance of the Utterance: assurance of Instinct; assurance of Experience; assurance of Form" (CP 8.374). Peirce distinguished propositions from assertions (MS 517) - a distinction also made by our contemporary logicians - but he drew some consequences that are not shared by his modem counterparts. For Peirce, the logical properties of a proposition are dependent upon its assertive function within the relationship between the utterer and the interpreter. It is true that Granger also refers to the capacity of communication which science reveals in its dependence upon the "couple utterer/receiver" (Granger 1979:21) [1]. But, as has been shown by Risto Hilpinen (Hilpinen 1983), Peirce defined the truth of a proposition "as the utterer's ability to defend it successfully against the interpreter's attack", and he analyzed quantifier phrases so as to give to "quantified sentences correct truth conditions" which approach the "modem game-theoretical interpretation of quantifiers" (Hilpinen 1983:268).

4 The Phenomenology of Knowledge: An Account of Reality

After having seen the logical presupposition of scientific truth in Peirce's logic, we may now want to explore the phenomenology of knowledge in Peirce's epistemology. In his "Prolegomena to an Apology for Pragmaticism" of 1905 (CP 4.530-572), Peirce defined a percept as the immediate object of our knowledge (CP 4.539); but it is not an immediate perception. A distinction must be made. Therefore, in "Why Study Logic" of 1902, Peirce distinguished "percept," "perceptual fact" and "reasoning." A percept is a sense evidence (CP 2.141-3); perceptual facts, upon which our inferences (CP 2.141) repose, describe percepts (CP 2.143): they are immediate interpretants of perceptual judgments (CP 4.539). True reasoning is different from percept and perceptual fact (CP 2.144): its conclusion is located between the remembrance of past percepts and past perceptual facts, and the expectation of a future (CP 2.145). The dynamic interpretant may be "Sympathetic, or Congruentive; Shocking or Percussive; Usual" (CP 8.370). The dynamic interpretant of a percept is a perceptual judgment which is a proposition of existence determined by its own dynamical object. And this latter is the percept (and the concept) that acquires the logical position of an abductive premise, which itself shows the perceptual judgment to be, an, albeit not necessary, abductive inference (CP 4.541).

To the perceptual universe are conjoined ever new universe sems, i.e., simple signs, which are interpretants of percepts, regarding "the Nature of the Influence of the Sign" (CP 8.373) And these interpretants of percepts open a way to Truth (CP 4.539), the highest abstraction. Besides sems, "simple signs," there are also phemes, or perceptual judgments, "with antecedent, consequent, and principle of sequence" (CP 8.373). Finally, there are delomes which replace arguments, and which represent change in thought-signs (Réthoré 1988:489). In relation to the higher abstraction, a false proposition would be a proposition from which could be deduced whatever a proposition itself opposed to a direct perceptual judgment (CP 2.327). In the consideration of perceptual judgments, the use of indexical expressions and predicates in these judgments also partake of the function of "Indices" and "Icons". Icons and Indices get their meaning within perceptual judgments. Perceptual cognition is thus not separable from a semeiotic expression of the percepts, because a sum of perceptual judgments exists through a flow of inferences.

So far, I have tried to expound Peirce's epistemology as depending upon a fundamental relation between the theory of knowledge and the theory of language, where the latter is understood as a logic and a pragmatics. This epistemology presupposes the mediation of mind between nature and society and a social principle rooted in logic. In this fundamental relation the pragmatic maxim plays its part as the nucleus holding together the different aspects belonging of Peirce's epistemology.

On the one hand, logic obeys the social principle, on the other, a natural phenomenology of mind leads to the belief and the rule and habit of thought as mental action. Moreover, it has been shown what are the logical presuppositions of knowledge in the proposition, which is always our representation, and it has been explained how the phenomenology of knowledge starts from the percept and ends with Truth. In both approaches to understanding Peirce's epistemology, the pragmatic maxim functioned as a sufficient principle: Peirce confirmed it like his other explanatory epistemological propositions.

By its capacity of insertion in a game-theoretical interpretation of quantifiers, as Hilpinen has shown, the originality of Peirce's epistemology is that it gives an account of reality and that it legalizes scientific communication.

5 The three Types of Inference

Peirce's analysis of the particularity of perceptual qualities and of the universality of abstract arguments, yielded three types of inference. These function as methodological connections in his epistemology. Deduction is the pure expression of thirdness, induction draws thirdness from secondness, and ab duction or retroduction draws thirdness from intuitive firstness.

Deduction and induction are well known and have been extensively explored from an epistemological perspective. But what is new and interesting in Peirce's epistemology is the notion of abduction. Peirce explains abduction in his "History of Science" of 1896 (CP 1.19-49), which was never completed. We must keep in mind that the first steps of scientific inquiry have been and are always difficult. Abduction is a preliminary knowledge or theory, which may be true or false, but which is always necessary to every first observation. "When a man desires ardently to know the truth, his first effort will be to imagine what that truth can be." In his early classifications of arguments, in 1878 and 1893, Peirce who was in search of a method (CP 2.372-388) called "hypothesis" what he was later to call "abduction." And he compared induction and hypothesis. Induction is the generalization from a number of cases of which something is true. Hypothesis is a sort of reasoning "where we find some very curious circumstance, which would be explained by the supposition that it was a case of a certain general rule, and thereupon adopt that supposition". Even though induction is seen as a strong kind of inference, which infers from one set of facts another set of similar facts, hypothesis is especially useful as inferring from facts of one kind to facts of another.

When we look at the epistemological fact that rules of induction and rules of deduction are now to be understood now as "canons of validation rather than of discovery" (Hempel 1966:18), we must appreciate the operation of abduction. Peirce would have rejected the classical and presumedly ideal conception of scientific inquiry, according to which such inquiry goes through four stages:

"(1) Observation and recording of all facts, (2) analysis and classification of these facts, (3) inductive derivation of generalizations from them, and (4) further testing of the generalizations." (Hempel 1966:31)

This is what Carl G. Hempel calls "the narrow inductivist conception of scientific inquiry" (Hempel 1966:31) and if we ask now the cognitive theory proposed by Einstein regarding his own discoveries, we find confirmation of this new perspective. For Einstein, the concept plays the part of a mental connection between sense experiences, but it is not identical with "the totality of sense impressions" referred to (Einstein 1954:291). Einstein assumed that a theoretical structure allows for theoretical descriptions that are not "directly dependent upon acts of empirical assertions" (Einstein 1949:674). In the same way as the Peircean abduction, concepts and system of concepts were seen by Einstein as a human creation. He presented the scientific, creative inference as going from the initial E (Experience) to the A (Axioms or principles), but through a J (Jump); with between A and the final E the necessary consequences: S, S', S" . Gerald Holton has recapitulated the complete cycle E-J-A-S-E as being Einstein's process of scientific theory construction, of which criteria are an "external validation" completed with an "inner perfection" (Holton 1986). It seems to me that the ways of Einstein's discovery is confirming the Peircean notion of abduction.

And in the Eight Lowell Conference of 1903, "How to Theorize" (CP 5.590-604), Peirce spoke of abduction as covering really "all the operations by which theories and conceptions are engendered". Criticizing Comte's theory of observation in the 28th Lesson of the Cours de philosophie positive (1830) (which, I guess, he misunderstood), Peirce explained what he understood properly with abduction: "any mode or degree of acceptance of a proposition as a truth". And he gave the truth condition of a good abduction, or of an explanatory hypothesis: "Any hypothesis [... ] may be admissible, in the absence of any special reasons to the contrary, provided it be capable of [...] verification." He concluded that was the doctrine of pragmaticism. But we can say that it is also the doctrine of scientific experimentation and theorizing.

We could add that for Peirce there would always be discrepancies between theory and observation (CP 1.132). Therefore, the scientist must constantly keep theories as flexible as possible to accommodate wayward data (CP 5:376): causes are not always precise. But it also happens that some truths cannot be supposed as not being universal. Necessity is then to be considered simply as a postulate: "a material fact which we are not entitled to assume as a premise, but the truth of which is requisite to the validity of an inference" (CP 6.41). The necessary modality is that of a habit as being an epistemic dimension of meaning (CP 8.376).

From the standpoint of the scientific inquiry to the conclusive abduction, Peirce has tempted to follow the real process of questioning which characterizes human mind.

University of Amiens


1 See (Granger 1979:14): "Une épistémologie du langage se développe dans la condition des sciences - celles-ci prennent les faits symboliques pour objets en vue de les décrire et de les construire." Compare this with (CP 5.9~119).


Almeder, Robert [1983]

Peirce on Meaning. In: Freeman E. (ed.) 1983a : 328-347.

Deledalle, Gérard (ed.) [1978]

Charles S. Peirce: Ecrits sur le signe. Rassemblés, traduits et commentés par Gérard Deledalle. Paris: Le Seuil

- [1979]

Théorie et pratique du signe. Paris: Payot.

- [1987]

Charles S. Peirce, phénoménologue et sémioticien. Amsterdam/Philadelphia: John Benjamin Publishing Company.

Einstein, Albert [1949]

Reply to Criticisms. In: Paul A. Schilpp (ed.), Albert Einstein: Philosopher-Scientist. Evanston, IL.: Open Court

- [1954]

Ideas and Opinions. New York: Crown

Freeman, Eugene (ed.) [1983a]

The Relevance of Charles Peirce. La Salle, IL.: The Monist Library of Philosophy.

- [1983b]

C.S. Peirce and Objectivity in Philosophy. In Freeman E. (ed.) 1983a : 59-77.

Granger, G.G. [1979]

Langages et épistémologie. Paris: Klincksieck.

Hempel, Carl G. [1966]

Philosophy of Natural Science. Foundations of Philosophy Series, edited by Elisabeth and Monroe Beardsley. Englewood Cliffs, NJ.: Prentice-Haîl.

Hilpinen, Risto [1983]

On C.S. Peirce's Theory of the Proposition: Peirce as a Precursor of GameTheoretîcal Semantics. In: Freeman E. (ed.) 1983a:264-270.

Hobbes, Thomas [1966a]

Opera philosophica quae latine scripsit. Aalen: Scientia Verlag.

- [1966b]

The Englisch Works. Aalen: Scientia Verlag.


Holton, Gerald [1986]

Einstein's model for constructing a Scientific Theory. In: Gerald Holton, The Advancement of Science, its Burdens. London: Cambridge University Press, 28-56

- [1988]

Thematic Origins of Scientific thought. Kepler to Einstein. Revised edition.Cambridge, MA.: Harvard University Press.

Kremer Marietti, Angèle [1982]

Entre le signe et l'histoire, L 'Anthropologie positiviste d'Auguste Comte. Paris: Méridiens Klincksieck. Paris : L’Harmattan, 1999.

- [1983a]

Le Concept de science positive, Ses tenants et ses aboutissants dans les structures anthropologiques du positivisme. Paris: Méridiens Klincksieck.

- [1983b]

Théorie du Signe: Comte entre Saussure et Peirce. In: Krisis, Summer 1983, International Circle for Research in Philosophy, Houston: Homage to Peirce.

- [1988]

Comte et la sémiotique. In: Recherches Sémiotiques/Semiotic Inquirv. Nos.1 & 2. Ottawa, 1988.

Paliaro, A [1950]

Logica y gramatica. In: Ricerche linguistiche. 1,1.

Réthoré, Joèlle [1988]

La linguistique sémiotique de Charles S. Peirce: Propositions

pour une grammaire phanéroscopique. Perpignan. Thèse de

Doctorat d'Etat. Exemplaire dactylographié.

Schneider, Herbert [1952]

Fourthness. In: Wiener and Young, Studies in the Philosophy of Charles S. Peirce. Cambridge, MA.: Harvard University Press, 209-214.

Strawson, P.F. [1974] Subject and Predicate in Logic and Grammar. London: Methuen & Co. Reprint 1976.