CHAPTER 1: Realism and Epistemology
Evolution of Epistemological Designs

"the common direction estimates reasonable to allot the majority of our feelings to causes which act of outside on our bodies. It does not believe only the room where we sat cease to exist as soon as we close the eyes or will lie down. It does not believe that our wife or our children is pure products of our imagination. In that, we can agree with the common direction; but where it is wrong, it is when it supposes that the objects resemble intrinsically perceptions which they cause."

Bertrand Russell

"All the question of reality, famous in philosophy, comes from the abusive value given to the word reality. If this me had been developed, and prevent from fleeing out of very thought clear, the problem had disappeared or had prodigiously changed."

Paul Valéry

Summary
At least in the Western thought, realism marks the starting point of the epistemological reflexion, and constitutes a reference for the later evolutions of this reflexion.

1. Analysis criticizes Realism. Does one have to burn Plato, Aristote and Saint Thomas d' Acquin? Spontaneous tendency of the human spirit, philosophical realism was formalized for the first time by the school of Milet, in pre-socratique Greece. It was then popularized through works of Plato and Aristote, then led until our days. It covers in fact three distinct affirrnations:

A) The World exists. There is no direction to discuss this assertion because that would question our own existence. But it is fundamental to stress that this indisputable realistic assertion, is independent of the two following ones.

B) The World is such as we perceive it. It is in fact a limiting attitude, never fully defended by the realistic philosophers who always admitted a minimum of subjective distortion. But criticism must go further and affirm that a priori, any cutting in distinct objects, any property noted in the World is marked by the characteristics of the observer, translated the relation between the observer and observed. The object of Science is to correct the effect of the perceptive distortion, allowing a continuous evolution towards a better comprehension of reality.

C) the ideas, the gasolines, the forms have a real existence and an independence. This point is more contestable. All the structures normally have a birth and a history, results from a creation by a system which precedes them and exceeds them. The concept, in particular, is a creation of the human mental activity, by abstraction of the data of lived. On the other hand, the concept once formed, can be detached from its origins, which gives the illusion of its character first and irreducible. The concept reflects one time the cognitive attitude, and there is no reason obligatorily to accept it like a reality "in oneself".

2. Oppositions to Realism. : Some grant the primacy to the environment, creating sensation the base of any knowledge and considering a progressive development of the subject knowing by pressure of the environment. With the opposite, in its native form, the initial idealism has especially the historical significance to have blamed realism, through the "revolution copernician": if many generations were mistaken while placing the ground in the center in the world, we must be wary today of all the obviousnesses. The other epistemological assertions of Kant are undoubtedly more debatable. The Kantian idealism re-examined by Helmholtz appears much more positive to us. Helmholtz affirms that the forms a priori sensitivity correspond to innate physiological; a sensory "screen" comes to fit between the external world and the conscience, outstanding by its own nature the perceived data of environment. It is implicitly the advertisement of the perceptivo-sensory interfaces, relayed recently by the discovery of the constitutional mechanisms of perceptive analysis.

3. Opposition of the scientific and philosophical approaches. It gradually took shape with XVIIème and XVIHème century in a climate of dispute of religious rationalism. The current positivist, inaugurated by Auguste Count, has especially a historical interest and of critical analysis of the other epistemological designs. On the other hand, phenomenology presents the highly positive character to center knowledge on the meeting between a subject and an object. But, fault of following a historical dimension which explains the genesis of cognitive structures, phenomenology is likely to derive towards realism from the natural species by postulating gasolines, forms, standards, under the aspect of irreducible and allowed structures a priori, jointly present in the Universe and the human understanding.

4. The New Alliance of philosophy and science.

4.1. G Bachelard and history of sciences : G Bachelard provides invaluable keys to include/understand the evolution of sciences. On the other hand, the author does not bring any data allowing to explain the ontogenetic cognitive development, to specify the mechanisms common to the ontogenetic development and the evolution of the scientific theories.

4.2. K Popper and conjectural knowledge : K Popper shows that induction cannot lead to the certainty and that any knowledge is only conjectural. It believes in an evolutionary epistemology where any new knowledge is a modification of a former knowledge, which in the final analysis supports any cognitive step on innate provisions. Popper believes in a "objective knowledge", created by the man but acquiring an existence independent of its creators, container of the checked theories but also of the false or unverifiable theories.

4.3. The constructivism piagétien : it was formulated by J Piaget in the continuation of J.M. Baldwin and E Claparède; it constitutes a coherent alternative to realism. Knowledge appears as the progressive organization of the data resulting from the meeting between the subject and its environment. The constructivism allows only a satisfactory confrontation between neurophysiological data and psychological data. One can however reproach J Piaget for not having made a sufficiently beautiful share with the innate cerebral organization. The constructivism finds before the letter the assertion of Niels Bohr, making of the man an actor and not only one spectator with the theatre of Nature, without falling for as much in the difficulties from the Kantian idealism. The discovery of the constitutional perceptive mechanisms brings a new dimension to the constructivism. The perceptivo-sensory interfaces provide an immediate interpretation of the environment, preceding any experiment, but marked by the characteristics of these interfaces. Later cognitive construction then seems a revision of these immediate data, established starting from the behavioral failures in contact with the environment.

4.4. Consequences of the analysis of the matter : the concept of substance is reduced to nothing by the atomic theory and quantum mechanics. Traditional realism made a broad loan with the difference in substances to explain disparities and properties in the Universe. The atomic theory is opposed to the irreducibility of the differences in substances. The comparison of discrete, concrete or virtual elements, is the true creative one of significances and new properties. Quantum mechanics develops the criticism of the substance and illustrates a conceived knowledge as the analysis of a meeting enters a knowing subject and a regularity of the environment.

On the whole, the historical succession of the epistemological designs grants a share increasing to the constitution and the human activity in the genesis of knowledge. But instead of being allowed a priori as in the Kantian idealism, the human activity must become itself object of study, in particular on the neurophysiological level and that of the ontogenetic development. The speech first must be questioned. The relation must occupy all the left place vacuum by a research without hope of the differences in "substances", which justifies a priori all the reductionnisms. Knowledge evolves to a continuous progress, but a progress which cannot be defined by advance. The theses of biological autonomy form part easily of this evolution of the ears temology.

We inherited our Master Piaget, the concern of presenting any new epistemological concept in reference with the traditional positions of various schools. The new ideas spontaneously nourish difficulties appeared at the time of the practical application of older ideas. New epistemological designs must thus be analyzed, to be well included/understood, compared to more traditional designs.

However, an exhaustive analysis of the various epistemological currents passed or present is out of our competence. We would more modestly wish to define the reference marks than we believed to be able to raise, specifying an evolution in the theses which have followed one another for twenty five centuries in Occident. In 1929, in "Process and Reality", A.H. Whitehead "general The safest characterization of the European philosophical tradition is that it consists of wrote has series of footnotes to Plato". This opinion appears to us justified overall, although outrageous. On the one hand, Plato is the heir to the school milésienne and on the other hand, the revolution copernician of Kant introduced indisputably a handing-over in question of many Platonic data. The fact remains that continuous Platonic realism to constitute the reference around which all the critical analysis in Epistemology is organized. The history of epistemology seemed to us to be able to be centered indeed on the initial formulation of realism, like wanted it Whitehead, then on criticisms to which this realism gave place; these criticisms appeared to us played a major part in the proposal and the construction for new epistemological designs.

1. Realism, spontaneous epistemological attitude.

A many doctrines philosophical, last or present, of the remainder often contradictory, could recommend realism. They have jointly to affirm the ontological existence of the objects of the surrounding world (Lalande):

- this existence is independent of the representations which the objects can provide on a subject expert; the object exists even if it is not perceived.

- but especially, the ontological properties of these objects are accessible, if not with perception, at least with the reflexion of this same subject. For realism, the subject perceives the object such as it "is".

- it joins the concept typically that the concepts which describe the existence and the properties of the object have a clean value and an existence.

- like J Piaget underlined, it results from it a certain primacy from the object on the subject, and thus a certain retreat in the importance of the subject, the acceptance of a reality independent of the spirit which conceives it.

1.1 Naive or vulgar Realism.

The least effort of introspection highlights than the contents of conscience correspond to one learned or one lived former, even if imagination makes it possible to exceed the limits of this lived. The learned speech, in particular, is in the conscious think tank. When the man starts to reflect on the direction of knowledge, which hardly occurs before adolescence, it already acquired a speech and a system of knowledge which it saw spontaneously, but that it cannot judge, for lack of a sufficient memorizing of the conditions of formation.

The first reflexion on this system of knowledge thus is spontaneously marked by immediate intuitions which rise in the way in which the relationship to outside is lived:

- there is an opposition between the me-subject and the object, each one having their independent existence.

- the speech whose origin is not perceived, appears the support of the thought, and it is developed a priori, in the allowed without proof and idealized concepts, as well as in the logical formalism.

Quite naturally, there is complete confusion between the object and the knowledge which one can have. This point of view is accentuated by the fact that knowledge before is very acquired near others and thus divided with him. Thus appears the naive realism which will be always an obligatory stage in the individual cognitive development, so much it seems obvious in first analysis and is essential before a reflexion does not show the weaknesses of them. Conversely, the history of epistemology could be brought back to long and difficult awakening of these weaknesses and to the need for correcting the realistic approach of the nature of knowledge.

1.2. The philosophical Realism of the Greek World.

What is worth in an individual cognitive evolution, is worth as much in the sociocultural evolution of the group and it is normal that the first formalized epistemological designs were realistic. C Hagège (077) proposes the very plausible assumption according to which a language writes phonetic constituted the essential reflexion of the oral language, to allow the analysis of it. The literal written translation at the very least supported the conservation of abstracted concepts, survivor with their creator and directing towards a clean life of the speech.

1.2. 1. The first formalization of Realism.

The Greek realism, initiated by the pre-socratique School of Milet and developed by Plato and Aristote, is a beginning of formalization of naive realism. Limited formalization which was supplemented gradually as criticisms obliged holding them of realism to perceive the implications and the weaknesses of them. On a point, Plato and Aristote meet, that of an anthropology which makes human spirit an irreducible reality, completely different from the material world, and isolated compared to him. On the other hand, Plato and Aristote are opposed on the nature of the ideas:

- for Plato and what was called later Platonic realism or the realism of natures, the ideas are independent of the objects which they are used to describe. The ideas are even more "real" that the individual and sensitive beings, which are only the reflection and the image (Lalande).

- for Aristote, separation between the idea and the object are artificial and the stress is laid on the operations of the speech. In addition, Aristote introduces another opposition, that on the one hand of the matter and on the other hand, the form which gives its individuality to the matter portion constituting an object.

1.2.2. Gasoline and Substance.

$$ Aristote tried to account for the permanence of the beings, corresponding to their ontological reality independently of the "accidents", variations of all types of the appearance of the beings. Aristote called upon the concept of whose commentators usually underlined the ambiguous direction. We will venture ourselves with saying that two concepts were drawn fromthe  aristotelician: - "the gasoline" which forms the bottom to be it, compared to the modifications which reach it only superficially or temporarily (Lalande) and which characterizes as well the completely passive beings as the active beings. The passage is easy gasoline to be it with the concept which describes this gasoline. - the "substance" translates permanence in a "subject", definite "in oneself" and "for oneself", which can act and undergo changes which leave it "the même"(Lalande). Enough paradoxically, it is the term of substance which was retained later on to define the chemically definite "matter" physically or. One can then retain certain points characterizing the aristotelism and all the doctrines which were born from it - the existence of "subjects" is marked. The subject is related to an irreducible principle, the substance. - gasolines and substances are irreducible. - they are the differences of gasolines or substances which basically support cognitive descriptions within a necessarily typological epistemology. 1.3. The quarrel of the Universals. At the third century after Christ, Plotin gave Plato to the honor, making loans with the incipient Christian doctrines in addition. One includes/understands easily the adhesion of Saint Augustin to this Neo-Platonism, adhesion which engaged practically the Western thought during the seven following centuries. It is undoubtedly the revival of a philosophical interest, very Net at Béranger de Tours in Xlème century, which explained the simultaneous appearance of the first detractors of realism. The dispute did not relate on the accessibility of reality to the analysis but to the realism of the ideas. Roscelin was the first nominalist: it refuted the general ideas which were for him only general signs. In opposition, the official teaching of the Church affirmed that the general ideas or universals exist independently of the things in which they appear. The conflict of nominalism and realism continued in the following centuries and supported a reflexion on the true direction of realism. That appears in particular in the evolution of the vocabulary. The terms even of "reales" and "nominal" were used as of Albert the Large one in XIIIème century, but the conceptualization of realism and nominalism was completely carried out only at the beginning of the XIXème century. This doctrinal opposition had two fundamental consequences: - it led for the first time to conceptualism, very positive doctrines formulated by Abélard at the beginning of XlIème century, and which wanted to be a compromise between realism and nominalism. According to these doctrines, the ideas or the universals do not exist in themselves, but they are not simple signs; they are clean operations of the thought. - the sterility of the first discussions scholastics has without any doubt supported in Xlllème century, the return to Aristote under the impulse of Saint Thomas d' Acquin, more opening being studied nature and the role of the feelings. 1.4. The Realism of modern Times and today. The "Cogito" of Descartes (046) marked a fundamental projection in epistemology by the taking into account of the subject expert. The Cartesian thesis does not remain about it less not marked Platonic realism in an ontological search for proof of existence. Following Descartes, realism remained extremely powerful in the modern thought in spite of the revolution copernician of Kant. It is in the center of the debate which opposed Einstein and the school of Copenhagen on the direction of quantum mechanics. The Platonic realism purest characterizes at the dawn of the XXème century the Bertrand Russell first manner (184) and the preexistence of the ideas is considered per G Bateson in 1979 the most fundamental discovery of Plato (O 10). The realism of Pierre Duhem (052), at the end of the XIXème century, much more is built and argued that the free assertion of Bateson. French physicist of the beginning of the century, Duhem was immediately famous abroad for his epistemological analyses. Its works are leaving to France the ghetto where had plunged the political and religious conservatism of their author. Duhem underlines separation complete and necessary between the scientific and metaphysical approach. On the other hand, it subjects the scientific theory to the facts, indicating that the theory is nothing other than a locus of coherence of the facts between them. Conversely it affirms the primacy of the fact, collected directly and in all independence of the theory. By doing this, it is made "the apostle of the direction common, only base of any certainty" what returns to a realistic profession of faith. Today, it is especially in the explanation of Mathematics that realism is most aggressive. K Gödel, famous to have shown the indecidability in the mathematical reasoning, was also a large theorist of mathematical realism; P. Benacceraf, H. Putnam, P. Maddy remain about it, following W.V.O. Quine, of burning defenders. It is remarkable that these authors are based only on logical arguments, therefore buckling on themselves, and neglect the knowledge obtained on the mental operation through the neurophysiology or of the psychological constructivism. In addition, several movements led during first half of the century, to postulate a speech first, without is formulated for as much an explicit reference to realism. Russian Pavlovian theories tried to place in a language defined a priori, all that conditioning could present of specificity at the social man compared to the animal. The linguistic structuralism of Jacobson, includes in a context very different the assertion from the speech first. Lacan took again the psychoanalytical explanations within the framework of linguistic structuralism. The development of Data processing directed towards similar conclusions. The logical operations being able to be easily reproduced by the machine, of many thinkers sought in the significances first, the characteristics of human mental operation. Through this historical permanence, it is possible to define a strict realism and its contents, under triple aspect: - realism affirms the existence of the Universe. - ontological realism affirms that the Universe, Nature, the environment, has properties "in oneself" independent of an observer and directly accessible to knowledge. The human spirit can have access to these ontological properties, not only in a direct and perceptive way but especially in a critical way on the basis of obvious truths which could not be seriously blamed; logic makes it possible to connect in a way constraining to these truths, initially hypothetical proposals. - irreducible primary structures, ideas or concepts of the speech but also forms or gasolines, have an existence and an eigenvalue, independent of the thinker who handles them. Thus space, time, the number, the good, the evil, the libido, the vintage and cook it have a clean existence, obviously supra-biological. It rises obligatorily from this last point which the subject which handles these abstract concepts is him also irreducible with a biological operation. 1.5. Critic of Realism. Implicitly or explicitly, realism was the object of many criticisms. The first formulations of Aristote ran up immediately against the objections of Pyrrhon and the skeptics who thought illusory of being able to establish a truth. Quite naturally realism was shaken by the questioning of data which seemed before obvious truths, like the geocentric organization of the Universe, or the belief which an object in displacement stops ineluctably if no force is applied to him. It is manifest that the calling into question of these truths involves a systematic question mark on all the other truths said obviously. Realism resisted nevertheless rather well criticisms, until appear, since the end of the XIXème century, of the objections much more serious. These objections could not be of a logical nature. Many authors, B Russell to H. Putnam discussed realism, in term of logical coherence. That it is a question of defending or of tackling realism, such a discussion hardly appears to us to have direction. It rests on a speech limited to a handling of concepts whose value is precisely the object of the debate. The realistic argumentation can thus nothing prove for an adversary of realism. The anti-realistic argumentation is also car-refuting since she criticizes the ontological significance of the concepts which she uses in her demonstrations. On the other hand, of the arguments more distant from a simple discussion of logic have a great value. 1.5.1 End of the truths first. Kant had extremely well understood that space and time constituted essential cognitive references, and the fact which applies to the Kantian idealism, is worth as much for realism. The Euclidean geometry, time and space absolutes of the Newtonian theories appear well to constitute obviously fundamental truths to which must be attached realism. However all the construction of these truths broke down at the end of the XIXème century. Leibnitz had already indicated that in its opinion, spaces and times were not other thing only relations between events. But in fact especially the analyses of Mach showed the inanity of time and space absolutes, independently even of the decisive experiment of Nlichelson-Morley. In parallel, work of Lobatchevski and Riemann relegated the Euclidean geometry to the row of a particular case among other possible geometries. On the whole the reference frames of space and time could not be regarded any more as points of anchoring of a realistic description. One could object that the relativistic reference frame espace/temps could be the true realistic reference frame. The argument is worth hardly bus much more still than Newtonian space and time, the relativistic espace/temps appears a construction, deduced by formalization from the experimentation. The number, another primary education reference of realism, appears reducible with a combination of operational activities, associating like it showed J Piaget and J.B. Grize (076), a cardinal and ordinal approach of the collections. The school constructivist of mathematics underlined the interest and the possibility of a reduction of a great number of mathematical concepts (002). 1.5.2 End of transcendent Logic. From 1915, date of the first publication of the theorem of Lowenhein, until 1936, date of the theorem of Church, several logicians showed (1 18): - either that certain fundamental systems to formalize mathematics contain necessarily indécidables proposals which can be neither derived, nor refuted. This assertion corresponds to the theorem of Gödel (139) but it is also close to the theorem of Church. - either that for any rather vast system, the concept of truth making it possible to associate each proposal of the system true "or" false "symbol a" cannot itself be formalized inside the system. It is the theorem of Tarski. In other words, logic as a whole comprises indécidables proposals and cannot formalize its own criteria of truth as soon as it is applied to an empirical situation not completely defined by advance. That by no means decreases the practical utility of logic but this one cannot show a clean, essential validity with the practical application of a realism without nuances in the analysis of the environment. 1.5.3. Diachronic analysis of the speech Platonic realism amounts affirming a speech first where the concepts have an existence independent of the objects which they describe. That should lead to a preexistence of the concept compared to the object which it describes. However, in fact, for the near total of the concepts used today, the process is chronologically opposite: 1' utilisaùon of the word and its contents appears secondary with the experiment. That is also manifest on the level dudéveloppement individual and the cultural level. - the child builds his semantic corpus gradually, learning how to name what he knew before. The fact is absolutely obvious for the first words, the little child inventing truly words which are clean for him to represent its experiments; Jean Piaget studied this phenomenon (149) well. We could note ourself the capacity of the young child to carry out incorrect verbal constructions starting from suffixes or of prefixes. Small a three year old girl announces that it "will descend" to indicate that it has a nausea. The invention of the verbal label used to represent an experiment is then manifest. The process is less obvious at the conceptual level but quite as real, in particular explaining the "witty remarks" of the children, who reflect simply an error on the exact direction of the words. The vocabulary grows rich according to precise and universal rules' for a given social group. When the word abstract and is learned before the conceptual significance of the word does not correspond to the cognitive capacities of the child, one can observe a later precision of the significance. Thus, we could note that the lazy word "is synonymous with malicious" for the child of quatre-cinq years. He designates the comrade "who cannot write" or "which is badly seen by the mistress" for the six year old child. It is only around seven years that "lazy" means "which does not want to read and write", and only around eight years that the significance "wants nothing to make" is acquired. Lastly, the distinction between idleness and idleness are even later. - whereas a seventeen year old pupil of our company can, according to G.A. Miller (130), to mean with precision average of eighty thousand different verbal forms, the primitive said current companies have a vocabulary limited to a few hundreds of words. In the language of our company, much of words translates a combination of words or concepts simpler, but that is also true for the primitive companies. It is thus necessary well that the majority of the concepts were historically created. It easily is noted remainder in the authors presocratic, often difficult to interpret because they are obliged to use concrete comparisons to label conceptual contents which we indicate today by specific abstract terms. Conversely, it is easy to note that many abstract terms are metaphors describing a concrete reality, like "the source" indicating the origin. - contrary, of many concepts, the unicorn to the phlogistique one, which a long time appeared to cover an undeniable concrete reality, appear today to correspond to empty classes. The comparative study of the foreign languages shows that they use conceptual cuttings which often overlap; it is then difficult to admit that the concepts preexist to this cutting. It is thus legitimate systematically to wonder about the contents of any concept. However the epistemological dynamics of realism is strongly devalued if in 1' utüisation of an abstracted concept, it is necessary to consider that it corresponds to an empty class. Realism is still put in charge when the categories like the color, the form, which seem particularly significant concepts by them same, appear as we will see it, to correspond in fact to unit cerebral mechanisms. 1.5.4. The realism rejected on the level of the concept could reappear on the level of the symbols. Since Freud and especially Jung, many anthropological studies show an at least relative universality of the symbols (051). There are few chances that this universality could be carried out by a universal diffusion following local creations and one could at least make the assumption d1une "patrimonial" origin of certain symbols, corresponding to a realistic point of view. This realistic explanation is however not obligatory, like it J Laplanche and J.B. Pontalis analyzing suggested the appearance of the primal phantasies freudiens, and obstructed with reason by a theory of genetic transmission hereditary of the symboles(1 17). They propose the alternative of an epigenetic explanation calling upon the uniformity of the contents of the infantile sexual life and the constancy of the later modifying influences. In an approach of autonomy, it would be necessary to associate there the concepts of homéorhésis * and créodes *, suggested by Waddington and taken again by Piaget (151). According to these concepts, a similarity even partial of constitution and close environmental conditions draw obligatory epigenetic ways or créodes. An autonomous force, the homéorhésis, brings back the epigenetic development * towards these créodes * when appear deviating effects. Without obligatorily following Laplanche and Pontalis on the validity of the primal phantasies, it is extremely possible to retain an explanation secondary, epigenetic, with the universality of the symbols and not there not to necessarily see a convincing argument in favour of the realism of natures. Ultimately, the fundamental pillars which seem essential with the assertion of realism, broke down and it is generally possible to explain "differently" the arguments which could be proposed in favour of realism. It becomes completely legitimate then to consider important corrections of traditional realism or failing this, to turn to nonrealistic epistemologies. 1.6 Corrections of Realism. They should be considered independently in the field of realism of the objects and on that of the concepts. 1.6.1 Ccritic Realism and the buckled Realism of B of Espagnat. The critic of the most radical realism and best argued comes from the philosophical physicists from quantum mechanics, Niels Bohr, Wemer Heisenberg and Erwin Schrödinger. Subjected to the requirements of thought which the approach of the microphysics required, these physicists understood that the conceptual revolution to which they were to yield, had a universal value. Schrödinger in particular (188), made go up at the school of Nfilet, Vlème century before Christ, which it named "the objectivation" and which it regarded as a deformation of the thought: Thalès, Anaximandre and Anaximène would have started the step which isolates the object from the subject which perceives it and a reality in oneself gives him. All realism was obviously built starting from this concept that Schrödinger condemned vigorously, affirming that the observer and observed forms an indissociable whole. Schrödinger defended the idea that the insulated object exists only according to the subject which thinks it. One can object that the school milésienne formalized only one attitude which existed spontaneously before but the influence on the development of realism was not reduced by it and criticism does not lose any therefore its value. It seems to us however that Schrödinger goes too far in its criticism because it does not take into account makes it become today obvious, that realism, outwards even with the realism of the ideas, covers two independent assertions concerning the surrounding objects: - the first relates to the existence of the world, of "something" corresponding at least partially to the object or the fact, existence affirmed without consideration of properties. Such a realistic assertion is limited to proclaim that the world, the object or the fact exist and of course that the environment is heterogeneous, that it has regularities, that it is possible to define properties in it, because failing this, the assertion of existence would not have a range. - the second relates to the relation between these regularities or these properties and knowledge. Traditional realism affirms not only that these regularities or properties have an ontological significance, independent of the characteristics of the observers, but also which they are easily accessible as such to the human understanding. It was the merit of G Santayana and of the American school of realism criticizes (Realism, in E.B.) to introduce this essential distinction. The analysis criticizes that we tried to lead door mainly on the second realistic assertion. To reject the first assertion completely would concern a belief not very reasonable solipsist and of which H. Putnam underlined the character of car-refutation (174). W Heisenberg (082) pointed out that the distinction established by Kant between ontological reality "nouménale" inaccessible to the human understanding and "phenomenal" appearance, translated an implicit belief in reality nouménale. W.V.O. Quine (176) was to further go that Santayana, affirming a reality of the things but disputing the fact that the structures which we allot to the things can correspond to real structures, external with us. Recently, B of Espagnat (062,063) arrived appreciably at the same conclusions as Santayana, Heisenberg or Quine. It showed that the thorough analysis of any object leads to an approach microphysics where the cognitive data relate strictly to the observer; the uncertainty and the relativity of the microscopic level refer on the object itself. B of Espagnat qualifies buckled realism this assertion of the existence of a world presenting of the regularities without possibility of analyses other than relating to the observer. We share this point of view which associates a recognition of the reality of the world and an epistemological design much more critical with respect to realism the direction full with the term, of a realism that H. Putnam qualifies the external one (174). It is possible to go further in the acceptance of the réalîsrne. Once allowed the distinction between noumene and the phenomenon, it is completely reasonable to admit that there can be a partial adequacy although nondemonstrable, between our representation of the authentic world and its regularities or heterogeneities. We even think that the development of knowledge precisely translates a progress of the level of adequacy. 1.6.2. The Lapse of memory and the Realism of the concepts. Realism criticizes or buckled realism appear answers adapted well to the contradictory problems posed by the significance of our representation of the world. The same reflexion can be led to the level of the Platonic realism which also covers two distinct designs: - traditional realism grants to the abstract beings an authentic existence, fully independent of the biological organizations which handle them. It is the extreme position of CH Hermite for example (quoted by Lalande) which did not fear to affirm in connection with the mathematical entities, "which they exist apart from us with the same character of need as the things of objective reality, and than meet we them or let us discover them, or let us study them like the physicists, the chemists and the zoologists." It was the position of Saint Thomas d' Acquin affirming that the universals had an existence "handle rem" in the divine spirit. - there is a design much more moderate which is satisfied to affirm than certain abstracted beings, often essential with the reasoning, appear indeed permanent, irreducible and independent of the mechanisms which call upon them as organizations which handle them, but without no assumption of principle being formulated concerning the clean existence of these abstract beings. This second point of view appears much more reasonable to us and can be capable to clarify the problems arising from the realism of the concepts, in particular mathematical realism. It is indeed possible to reconcile the independence and the irreducibility of the concept with a built origin, bringing back to be to it abstract to the result of activity of abstraction carried out by an organization, but being then fully detached of this organization and the objects from which the abstraction was derived. This design of the only irreducible concept seemingly, returns to two concepts: - the first concept is that of the regular emergence of original structures, to start from elements and the play of an organization, but being released then from its origins; only a deliberated historical analysis can sometimes find the conditions of emergence but the logical reflexion can build a probable model of emergence. This thesis of emergence is one of the ideas key which we try to defend in this work, - the second concept is that of the effects of the lapse of memory. The individual or collective memory well more perennializes the final result of a constructive process that the circumstances of construction. In addition, the mechanisms manufacturers do not have no need to be well analyzed so that the result of their activity are the subject of a mnemic fixing. A concept can thus be created and its significance selected while at the same time the circumstances of its creation are forgotten. However the majority of the concepts which we use testify to this selective fixing: - the history very often enables us to show that a new concept was formed, without for all this the conditions of formation are well defined. - we use constantly and without too much embarrassment, of the concepts, while having completely forgotten the circumstances during which we learned these concepts. Ultimately, it is enough to consider that the concept once built, is detached from the conditions of its genesis so that it takes the appearance of a structure first irreducible. A comparison could be made with the biological organizations. If we accept the transformism of the species, which is the case of almost all the thinkers today, we must consider that any current organization is the result of a history. However, we can approach the reality of this organization in almost total independence of this history. In Epistemology, the situation is partially identical, partially different. a) it is manifest that all the serious arguments of realistic are based indeed on the independence and the permanence of the concepts, their nonat least apparent reductibility. It is often completely reasonable to accept this point of view, while retaining that it is reconcilable with one or the other of the two hypotheses adopted higher. Independence can translate a clean existence or a mental construction whose mechanism was forgotten. As J Piaget underlines it, only the observation of the young child, possibly associated with the study with the primitive human societies, finding the conditions of formation, can bring to the debate, of the valid arguments. b) on the other hand, the use of the concept in the speech, appears very different according to whether one or the other of the two assumptions is retained: - to suppose that the concept has an authentic clean existence, it is to conclude that it has invariant properties, that we can progress in the knowledge of these properties but that they are not modified by it for all that. It becomes legitimate indefinitely to continue a reasoning established on the only properties of the concepts used, in the condition however that the correspondence between concept and word is bijective and spontaneous. - to suppose that the concept is the product of a former human reflexion in circumstances forgotten since, resulted in thinking that the concept cannot really have final invariances. Under these conditions, the use of a concept in a circular speech seeking to lead until the end the consequences of invariances selected, is likely all to lead to the error. For our part, it seems to us completely legitimate to commonly use the concepts in their grateful independence and permanence. On the other hand, refusing a clean existence to them, we refuse that a speech led until the limits of the current definition of a concept, can be based absolutely on this definition. We will see further how G Bachelard could show how much the strict definition concepts was the cause of errors. 1.6.3. Mathematical Realism vis-a-vis in Constructivisme. The opposition between an authentic Platonic realism and the acceptance of a relative irreducibility of the concepts are particularly important on the mathematical level, because it is the field where the realism of natures is most active today. It appears to us that the second described assumption should satisfy realistic, formal and constructivists. The attitude of the mathematical constructivism which would like not to use that mathematical objects completion definite is perhaps épistémologiquement healthy but it is illusory bus in the absence of a complete historical rebuilding, the constructivist always needs to call upon nonreducible concepts to establish a definition. Conversely, the only fact of affirming that a many mathematical concepts must be used like irreducible entities, without obligation to be concerned with their origin, should answer all the requirements of realistic, making the economy of the many difficulties related to a strict platonism. Let us begin again on this point, the most recent thesis of the realistic design of the number such that P. Maddy considers it (123). One can of the remainder stress that if P. Maddy is declared realistic, explanations that it borrows from Hebb locates it well close to conceptualism. Paradoxically, P. Maddy calls some with work of J Piaget and A. Szeminska(161) stressing that the relations between subsets cannot be correctly perceived before the properties of the number. It seeks to show that the theoretical notion of the unit is a realistic data and that it is indissociable concept of a number: "the knowledge of the number is the knowledge of the unit because the number is a property of the unit. Reciprocally, the knowledge of the unit presupposes the knowledge of the number." We notice that P. Maddy, like K Gödel, share of the report which it must exist "some thing" between the simple combination of feelings in contact with a concrete collection and the concept of unit (quoted by P. Maddy). All the perceptive organization, description after initial work of Hubel and Wiesel (V-C) easily explain in fact this "something", and that much better than the designs of Hebb to which P. Maddy refers. The author moves away from the proper explanations of Piaget to which it refers, simply saying Piaget that the intensive quantification of the subsets presents the same requirements as the construction of the number. The analyses of P. Maddy lead us to the following reflexions: - it is indeed true in practice that the quantified comparison of subsets is supported considerably by the handling of the numbers. - weak extent of the field of conscience, basic concept on which we will return, prevents from perceiving the number simultaneously, at the same time like an operational tool in the comparison of the subsets and at the same time, like a construction learned, by fusion of a class and an asymmetrical relation, according to the diagram piagétien. So that the number is correctly used, it is necessary thus that it is actually detached from the circumstances of its creation. - the detachment of the number compared to the circumstances of its training make of it a tool which appears first and which is actually used as irreducible structure, but that does not contradict of anything a former process of construction number. Ultimately, it can be desirable to regard mathematical concepts as first. The restrictive properties of the conscience even require "to temporarily forget" that the concepts are built but on the other hand, mathematics does not have probably anything to gain and very to lose in the assertion of a clean existence of mathematical objects because they would cut all data then that the genetic psychology and the recent neuropsychological analyses can bring explanatory value to them. 1.6.4. Internal realism and conceptualism. We have to insist much on the realism of mathematical concepts, in particular in the Anglo-Saxon authors, because it appears exemplary to us to locate the internal realism of Putnam compared to conceptualism: - perhaps by bad luck, we hardly met the term of conceptualism in the Anglo-Saxon philosophical literature. Encyclopaedia Britannica referred there not even in an article largely documented on Abélard. We wonder then if "realism interns" of Putnam is not in fact very close to a conceptualism nativist which would affirm the existence in the mental activity, of "forms" or of "mechanisms" which would not have to be learned, would thus be registered in the human constitution but would not correspond for as much to entities of the external world. - but, we postulate ourself an innate organization more developed much than J Piaget indicated it. However we have the concern of connecting any innate capacity to a neurophysiological analysis which highlights it. In addition, we preserve the distinction necessary between a constitutional structure and the mobilizable design which can of it be derived. Confrontation between a realism "interns" and a modified constructivism must be placed on the only plans of the neurophysiological analysis, of the observation of the young child and a theory of the formation of the mobilizable designs. The question which arises appears us then the following one: such concepts, of the remainder more operational than descriptive, like the opposition between the figure and the bottom, the element and the collection, are first, included in the cerebral constitution, and immediately usable? Are they on the contrary quickly and precociously derived from more fundamental concepts which would correspond as for them, to constitutional structures? The question appears completely comparable to us with that which relates to the nature of grammars, and which was the object of a conflict between N Chomsky and J Piaget (162). It does not appear possible to us that is given today a clear answer, and in the current state of knowledge, it seems to us of little importance to choose one or the other of the answers. - we would have despite everything tendency to think that the forms or mechanisms supposed first according to the diagram of internal realism, are designed like too mobilizable in the mental activity being authentically first. A time of exercise in contact with the environment and reflective abstraction seems essential to us to ensure a mobilization. Conversely and contrary to J Piaget, we think indeed that presumedly constitutional forms and mechanisms within the framework of internal realism, are much closer to constitutional cerebral mechanisms than J Piaget thought it. The distinction between realism interns and a renewed constructivism becomes a question of nuances then. 1.7. Relative, minimal or balanced Realism. The analysis of Realism thus leads us to think that the contradictory opinions which are carried about it, raise of a mixture from point of view of which the ones remain essential, the others are not only contingent but harmful. - it would be aberrant to dispute the reality of the world but it is extremely criticizable to affirm that it is presented spontaneously at any observer under properties nouménales. "reality in oneself" is a potentiality independent of any observer and any observation, so that qualifications emergent if observations by some observer produisent(M. Mugur-Schachter), something which admits the single qualification to be qualifiable, independently of the fact so yes or not of the qualifiers exist and agissent(C. Castoriadis) (138). It appears essential to us to add that this something is located in a space at the same time not limited and heterogeneous, which distinguishes it from a simple continuous space of probability. This heterogeneity * must be marked but it cannot be analyzed because it is completely masked by the cognitive action which begin with a subjective cutting, creating a heterogeneity * other. All "object" of analysis of the external world to which knowledge relates is indeed the result of a preliminary subjective step of division reality. Conversely, this reality is essential as W Heisenberg underlines it: "One cannot admit all to reduce to the act to observe; in his science, the physicist must postulate that it studies a world which it did not manufacture itself and who would be present, primarily unchanged, if the scientist were itself absent (083)". - mental operation requires the use of general ideas, forms, standards or mathematical concepts, handled as of the structures first but ü is completely allowed and even desirable to attach these mental structures to circumstances of old construction from which they were then detached. In addition, we correct ourself the constructivism piagétien while recognizing, at least in the field of the perceptive organization, the "forms" and especially of the "mechanisms" registered in the cerebral constitution and immediately usable. The whole of these steps corresponds to the "balanced realism" of G Edelman (058), minimal assumption only postulated but essential to give a significance to any reflexion on knowledge. - we think that a minimum of reflective abstraction is essential to pass from a constitutional mechanism to a mobilizable design in the mental activity, but conversely that a constitutional mechanism can be extremely close to a concept appearing almost first in the mental activity. Through these reflexions, the differences between a minimal realism and the constructivism are very attenuated or even cancelled, which could perhaps decreased the interest of a search for nonrealistic epistemologies. 2. Traditional nonrealistic Epistemologies. They mark the primacy of the external world with sensualism and empiricism, or the primacy of the subject expert with the Kantian idealism. 2.1. Sensualism. It is usual to describe under this term an active current of thought of Epicure with Ernst Mach, which privileges the feeling in the cognitive process. Aristote had already attenuated the rationalism of Plato by affirming the role of the feeling which assimilates the external forms to allow the analysis of it. Some thirty years later, Epicure reverses the roles and creates sensation the time first of knowledge. It does not neglect therefore the action of the subject through "the anticipation" which immediately grants a significance to the feeling, and it often speaks about "reason" and "raisonnement'. In XVIlème century, the opposition of sensualism and rationalism are found in the controversy which opposes Gassendi and Descartes. A little later Condillac radicalizes truly sensualism. More recently, the primacy of the feeling, affirmed per Mach, appears one of the fundamental characteristics of logical empiricism. Generally, the theses épicuriennes had the great merit to mark the limits of rationalism. However, it was difficult to see in the different feeling thing that a translation passivates "reality", which made quite difficult an explanation of the development active of knowledge. As we will see it (V-C), the discovery of a double system of analysis of the environment upstream of the conscious reflexion, completely renews the role of the feelings in knowledge. 2.2. Empiricism. By need, empiricism recognizes the importance of the feeling but its originality is elsewhere. As the majority of the philosophical doctrines, empiricism is defined as much in opposition to other doctrines as by itself. It is opposed to the rationalism inneist which grants innate mental properties to the individuals, as Descartes or Kant does it for example. Empiricism condemns the mechanisms and the innate ideas but moreover, it minimizes the mental activities of the subjects acquiring of knowledge. Robert Bacon was, at the end of Xlllème century, the precursor of empiricism. He stated to prefer the observation with the deduction in a time when the triumph of the scholastic supported a reflexion in closed loop. Francis Bacon, at the end of XVIème century, is regarded as the father of English empiricism (E.B.); he encouraged with the collection first data of experiment and observation, affirming that the arrangement of these data is established then systematically, quasi mechanically. John Locke, at the end of XVIlème century, was especially critical innate knowledge: the child does not know anything. David Hume, at the XVIIIème century, tried to show that it is impossible to affirm the bonds of cause for purpose; one can only note that a state "has", apparently causal, regularly precedes occurred by a state "B". The range of the reasoning is thus very reduced. The conditions were then met for a systematization of empiricism but the political ideology played certainly a complementary part. It is with a number of ulterior motives protestors that Diderot took again the thesis of Locke of the close-cropped table: the brain is initially only one slate without inscription and the child must all learn. Thus the superiority can be refused a priori certain social categories. By doing this, Diderot was not fully aware that its thesis related simultaneously to learned knowledge and the mental structuring but it announced nevertheless the introduction of a factor time and a secondary structuring of the individual. With the end of the XVIIIème century, empiricism moved towards the Natural science. Lamarck postulated the transformation of the species but it remained ambiguous on a fundamental point: are the transformations dependent on the activity of the subject with respect to a medium or on an instructive action of this medium? With the second reading of the author, it seems well that Lamarck supported the first thesis to explain the animal transformism and the second to translate the vegetable transformism (075). It is thus to James Hutton that it is necessary to grant the first creed fundamental empirist, formulated within the framework of geology and the geography: the structures present, mountains, valleys, lake, are always currently the existing effect of the continuous and very prolonged action external forces (E.B.). This displacement of empiricism towards the natural science allowed one century later, an orientation towards the search for experimental bases in the field of psychology, initially reserved for philosophy. Pavlov, at the end of the XIXème century, paradoxically itself little convinced of the value of explanation of the conditioned reflexes in the higher mental processes, founded the Russian school of reflexology. Watson took again reflexology in the United States, melting the behaviorism, developed during the century present by Skinner and Lashley. - starting from a corpus of innate reflexes, directly not very effective, the repeated presence of regularities in the environment reveals derived reflexes, known as "conditioned", which allow an adaptation of much better quality the medium actually met. In fact thus the regularities of environment constitute the motive fluid of the appearance of the learned conduits, the organization being hardly more than the locus where the effects of the regularities concentrate. Skinner slightly increased the role of the subject by describing the operative conditioned reflex, where a spontaneous action of the organization is reinforced by a reward. All in all, according to the diagram behaviorist, the social organization is obviously first compared to the individual organization and the human behavior is nothing other than a sum of learned practices. - to try to supplement the neurophysiological support of the conditioning and in spite of former work of Broca, Lashley postulated and tried to show the equipotentiality of the cerebral systems in experiments: all the cerebral surfaces have an identical function at least initially and can compensate the ones the others; the brain by itself is unorganized. With the test of the facts, empiricism was a double failure - in 1937, Konrad Lorentz could brocarder doctrines which had the ambition to explain the animal conduits in the natural environment and which was unable to envisage the behavior of the white rat in the given medium of the laboratory. - in 1945, Lashley itself recognized that the thesis of the cerebral equipotentiality was insupportable. However, this failure of empiricism was positive. It allowed in contrast the assertion of the school of ethology objectivist of K Lorentz and N Tinbergen, which showed the existence of the innate determinants of the behavior. In its turn, same excesses of this school (177) supported at T C Schneirla and J.S. Rosenblatt, the thesis of the animal trainings without parental models which has obviously a significance of an exceptional interest for the theses of biological autonomy (085,177). 2.3 Kantian Idealism. The inertia of the acquired ideas made that it ran out more than three hundred years before a philosopher does not take the full epistemological measurement of discovered with Copernic. Kant, the first, pointed out the importance of a data which had been recognized but not commented on: the fact that the sun turns around the ground, erroneous and translating an illusion of our directions, had been accepted however without discussion since the paddle of humanity. If many successive generations had remained in the error on this precise point, it became obvious that many beliefs appearing to correspond to obviousnesses, could be quite as illusory. Kant proposed a generalization starting from this report: we can know only what is recognizable for us and our recognizable is in conformity with our faculty to know (012). So knowledge "in oneself" becomes a knowledge "for us". There is in the analyses of Kant, two essential and relatively independent points - the objects have many ontological properties, corresponding to a "noumene" but this noumene is inaccessible to the human understanding. On the other hand, the same objects have for us an appearance and constitute "phenomena" accessible to the cognitive analysis. Knowledge thus relates exclusively to this appearance. - the meeting with the object takes place in space and time; it is necessary thus that space and times precede the experiment: they are the forms a priori of our sensitivity. Knowledge cannot be immediate because it would be just as it is a priori, independently of the object in oneself which causes it. Knowledge synthetic, is thus built. Noting Cogito of Descartes becomes constituent Cogito. Knowledge translates our cognitive faculties into action. The distinction between ontological reality and accessibility with this reality is probably one of the most fundamental conquests of philosophy and covers any epistemological design. The second point of the Kantian doctrines poses more problem. The demonstration of the character a priori of space and time seems to us a philosophical play which would have value only in so far as the concepts used have an absolute value, making it possible to lead a reasoning to its completion. The Kantianism returns in particular to an absolute and debatable definition of space and time. On the other hand, the assertion of the character synthetic and built knowledge is much more positive. Though it is, the Kantian idealism marked an essential time in the history of philosophy. That explains its prolongations and its influence, in particular in Germany until the end of the XIXème century and even later in France with L Brunschvicg (030). This influence is found in the various neo-kantian movements. 2.4 Neo-Kantianism of H. von Helmholtz. Helmholtz was before a whole neurophysiologist and one of largest. But studying the sensory processes, it was led to read again the work of Kant in an original way. Helmholtz refused the direct, passive assimilation and automatic of the surrounding world. The feelings for him, at the same time body modifications and one are lived of subjectivity, as much built than given; in this direction, Helmholtz followed the teaching of Kant well. On the other hand, the access to the conscience is not immediate for Helmholtz, because between the excitation and interpretation, there are active receiving bodies which mean the feelings. They are these receiving bodies and not the conscience which work out time and space. The forms a priori of the sensitivity are thus preserved but become properties of the physiological clean body, to which the conscience does not have access, at least immediately. To some extent, a priori becomes innate physiological (H.P.). This reinterpretation of Kant is essential because it puts side the major difficulty of the immediate apprehension conscious of time and space. Helmholtz was not satisfied with an assertion but began a demonstration and it is with reason that P. Buser and Mr. Imbert could dedicate to Hehnholtz their recent publications on the functional neurophysiology of the vision and hearing (033). Helmholtz promulgated the trichromatic theory of the vision which was confirmed one century later. It also affirmed the need for specific resonators auditive for each auditive frequency sound, which was also checked. The work of Helmholtz is thus capital by opening the field of a subjective activity authenticates different from the conscience and forming an obligatory screen between the subject and the object. 3. Opposition of the scientific and philosophical approaches. The Greek or Roman thinkers of Antiquity did not oppose science and philosophy. The same report can be made concerning the Arab thought and the Western thought at least until XVIlème century. The eclipse during the Average Western Age of a study of nature is related to social conditions privileging theology much more than with a doctrinal refusal. The situation changed then radically, probably because of the opposition between the currents of religious rationalism and those of the experimental study. Separation, started once, continued on its own impetus. The current positivist is thus marked by an exclusion or a negation of metaphysics. The spiritualism of Bergson or the phenomenology of Husserl is conversely built to some extent against Science, or at least against certain aspects of Science and apart from it. The effects of these oppositions on epistemology are manifest. 3.1. The current positivist. He was born with the Course from positive philosophy from Auguste Count, but really appeared in a coherent way much later. 3.1.1 The Positivism of Auguste COUNT. Few authors emitted opinions or interdicts which more quickly and were more completely contradicted by the evolution of the ideas, that Auguste Count. Michel Serres mentions the fact (189) and explains it by stressing that Auguste Count, alive in first half of the XIXème century, is in fact the ultimate heir to the current encyclopedist of the XVIIIème century. In spite of this failure, Auguste Count was the initiator of an essential epistemological movement which continues nowadays, current that Lalande summarizes as follows: - only the knowledge of the "facts" is fertile. - the certainty is provided by the applied sciences. - the human spirit can avoid the error and verbalism, only while being unceasingly held in contact with the experiment and while renonçant with all a priori. - the field of the things in oneself is inaccessible, the thought can reach only relations and laws. The considerable development of sciences, inclusion in the scientific thought of reflexions on space and time, traditionally reserved for philosophy, explain the importance of a current positivist remained very dynamic. One notes in fact a certain number of orientations more or less connected the ones to the others, and apart from the circle of Vienna, credit of 1924 to 1930, it is relatively difficult to systematize definite schools. The current contemporary positivist took the names of logical positivism and logical empiricism (095). One can find three inspirations there. 3.1.2. The inspiration empirist. Independently of scientific empiricism prevailing in psychology, a thought empirist of distinct origin is found in a logical and epistemological approach. That is undoubtedly related to the very strong personality of E. Mach and with its influence on the members of the circle of Vienna. E Mach affirmed that an object is summarized with the whole of the feelings which it can cause at the observer. The major disadvantage is a return to Francis Bacon, without interrogation on the mechanisms of assimilation of the significant data since this assimilation is regarded as self-explanatory. In fact, empiricism was especially called upon by logical positivism to condemn all a priori idealistic, any phenomenologic gasoline *. The ground was thus released for a logical and formal approach. A special mention must however be made analyses of Mr. Schlick, one of the founders of the circle of Vienna. It was very conscious of the irreducible character of the subjective feelings in their qualitative aspect and of the difficulties of the intersubjective transmission of these aspects. It has of this marked fact that the aspects configuratifs drawn by elementary feelings in the same way standard were much more important than the qualitative type of the feelings (quoted by L Rougier, 183). We will find the validity from this point of view (V-B). 3.1.3. Science language well done. Mathematical logic was born with Leibnitz then it calmed down. Logic was to reappear at the time of Auguste Count with B Bolzano and especially with the publication into 1847 of "Mathematical analysis of Logic" of G Boole. In this work and the following, Boole invented a system symbolic system allowing to treat logic like the algebra and supporting considerably the work of his successors. From 1879 to 1903, Frege tried a logification of mathematics which had a very important repercussion in all the logicians. These are the this work which allowed the publication by Whitehead and Russell, of 1910 to 1913, the true nap that constitute Principia Mathematica, providing an essential tool for analysis of the language. The junction of this logical work and the empiricism preached per Mach leads to the synthesis carried out by Mr. Schlick and his colleagues of the Circle of Vienna in 1924. Mr. Schlick integrated logic but it refused the realistic character of it and saw there the only rules of a tautological language whose conclusions are latent in the premises. Logic regularizes the language in the expression of the experimental truthes: science is brought back to a language well done bearing on the significant data. From this concept, Mr. Schlick and especially R. Carnap (E.B., Positivism and Logical Empirism) will try to show the importance to clarify the direction of the basic concepts and the assertions, in particular in the field of sciences. It is the only means to avoid the pseudo-problems and the use of empty concepts. In addition, the authors find absurd to tackle the questions without response of ontology. This is why they are not taken any as K Popper with the aspects pseudo-scientists of the Marxism or the psychoanalysis freudienne but with the metaphysicians of their time, in particular Mr. Heidegger. Using all the data of the modern logic, in particular work of Russell, Schlick and Carnap endeavour to show that metaphysics largely calls upon concepts badly definite and resulted in putting forth proposals logically incorrect or stripped of direction. It should well be recognized that criticisms of the positivists of the circle of Vienna are partly founded and one can sometimes wonder whether the metaphysicians often deliberately do not seek the blur to avoid contradiction. The severe remarks of L Rougier, defending logical empiricism, perhaps are not completely stripped of bases: "the majority of the metaphysicians are like the Pythea of Delphes. They are useful of language so impenetrable that, when you show to them that it is absurd, they have always the loophole to say to you that you badly heard them. To defend their position, the metaphysicians, as the omens of antiquity, are intended to support the right to be useful of nondefinite concepts, fluid concepts and in a state of perpetual metamorphoses. It is like if, playing the Failures, one allowed oneself, without informing the adversary of it, to modify the rules constantly of them, or as if one made physics while modifying continuously, without informing of it the public, standards of mesure."(183) But semantic hygiene has the narrow limiting ones. If Platonic realism is rejected, the concepts have nothing any more but one relative significance which is not only opened with the revisions but also partly unspecified; so no logical reasoning can be led to ultimate conclusions. Moreover, the concepts are implicitly marked by the theories which they describe, which causes of the loops indécidables. Hertz had shown hundred years ago already that the coherence * of the Newtonian theories came from what concepts of mass, time, of space, used by Newton were defined according to these theories. Hanson, Kuhn and Feyerabend (095) pointed out that the real significance of the most descriptive vocabulary, apparently most observational depends on the theoretical context in which it is used. Logical formalization has itself, we saw it, of strict limits of validity. On the whole, the proposals most correctly built can have only one very relative value. Conversely, as we will see it, the benefit of semantic hygiene is itself relative because a correct transmission of information can be completely carried out in the presence of noise, like J.D. Cowan and S. Winograd showed it (225). 3.1.4. Criteria of checking of the empirical direction. The proposals can have an empirical direction or a formal direction. For the school of logical positivism, an expression which contains logical terms and empirical terms, is empty of empirical contents if its truth or its falseness depends only on the rules of use of the logical terms. The only proposals which have an empirical direction are those which translate an observable fact being able to be checked or cancelled concretely. In practice, this assertion of the criteria of checking poses many problems. What can one think for example of the empirical direction of the sentence "All the cats have claws" which can be checked only on one limited number of cats. The principal merit of the position positivist was to cause the critical analysis of K Popper. This one, finding L Brunschvicg and the criticism of the nonrefutable theories, showed that the empirical criterion of validity was to be given up and replaced by the criterion of refutability: a whole of assertions has scientific value only if it is possible to consider facts being able to result in refuting these assertions. New theories, and they is very well like that, are often conjectures dared and far from the tradition. It is thus essential that these theories are subjected to criticism, only one established empirical contradiction good having to suffice for a refutation. The development of knowledge must consist in correcting the errors by refutation from the existing theories, either which one discovers that they are logically nonconsistent, or especially that they are cancelled by facts. Ultimately, the movement positivist did not carry out the epistemological revolutions directly that hoped for its promoter. It had however a very important indirect influence by the reactions which it caused. It showed the defects of a science which would like all to include/understand and the vanity of what Michel Foucault called "the épistémé scientific one" of modern times. An opposite criticism could be addressed to phenomenology, when it appears artificially detached from the scientific data. 3.2. Phenomenology and Existentialisme. Historically, phenomenology was born in 1891 with R. Husserl. He was a mathematician and without much training in psychology, which explains its very different attitude with respect to mathematics and of psychology (091). Later on and although Husserl had a persistent influence, numbers its continuators or pupils introduced from the original points of view, in very first Heidegger place. The goal of the phenomenologic approach seems an effort to define "gasolines" or standards, inseparable from the facts or phenomena but which are not immediately accessible to the clear conscience. These gasolines and these standards are also present at the subject expert which has the intuition of it. For L Rougier, the completion of phenomenologic knowledge is a mapping of the gasolines, standards, forms or structures irreducible in their double reality, objective and subjective (183). In the field of positive sciences, phenomenology does not have a conflict position because it affirms that the scientists resort, without suspecting it, with the intuition of the gasolines * (152). The problem is very different on the psychological level where the psychologism, which aims in fact especially the theses associationists, is highly criticized. The phenomenologic method allows only a "reduction", freeing the subject from spontaneous limitations caused by the former experiments, enabling him to find the gasolines, while being discovered itself. To some extent, the subject learns that it is not only one part of the Universe but the base of the apprehension of this Universe. Heidegger introduced a new approach which is a redefinition of the verb "being", announcing the idea sartrienne of the man which is only what it makes, the existence preceding the gasoline. One of the originalities of phenomenology comes from the postulate of a fundamental interaction between subject and object, therefore of a starting point of the knowledge which is neither in ontological properties of the object, nor in a purely synthetic activity of the subject but well in the relation sujet/objet. Phenomenology does not say "I think" but "I think of something", announcing the assertions of J Schlanger for which any knowledge is a knowledge of something by somebody (186,187). Conversely, the object is not object in oneself but primarily perceived object. Neither Husserl, nor Heidegger really sought to found knowledge. They rather thought of managing by their methods to specify standards allowing a redefinition of the world (152). The fact remains that the phenomenologic position appears very close to us by many sides to the theses which we try to support: - perhaps initial "nothing" of the existentialism translates in a a little complicated way, the very simple assertion of a stage first in the biological life of the individual, where the conscience yet really does not exist as such, because it is reducible with a pure neurological mechanism, without mental contents. - the starting point of learned knowledge is a relation between a subject and an object of knowledge. In this direction, any individual must build itself all learned knowledge that it can acquire, and consequently contents of its own conscience. - the knowledge learned translated a redefinition from the world starting from the relation sujet/objet, as well from the representation of oneself of that of the Universe. the sentence of Sartre "the man is what it becomes" could be slightly modified and to correspond to our thesis the man does not acquire the conscience and the knowledge of what it is and of what it meets only while acting. These very rich ideas of phenomenology, our opinion, are unfortunately associated positions which seem to us criticizable: - phenomenology marks a return to the realism of the natural species. One as well sees it through gasolines as in the "good forms" of the gestalttheorie *, psychological movement very near to phenomenology. - phenomenology and existentialism historically accentuated a cut with the scientific data, in particular those which can be collected on the knowing conscience or the nature of the perceptive relationship to the environment. Phenomenology thus appears to us to necessarily introduce a dualism * which opposes an irreducible conscience and the biological constitution. - no place is left with the development of knowledge by stages, however obvious as well in the cultural evolution in the ontogenetic development. So Piaget could say (152) that phenomenology and the existentialism remained on a plan of definitions of standards and could not claim with the universality on the epistemological level. That is not too awkward in the context of a purely philosophical for "standard" but crippling search in the application to the fact. Two examples which we lived personally appear very suggestive us 3.2. 1. Phenomenology and perception. The gestalttheorie *, near to phenomenology, was interested much in the law of constancy which explains why we grant the same size to the objects, whatever the distance to which they are compared to our eye; the size of the retinal image however varies with the distance. The gestalttheorie * saw in this law, the result of an irreducible total organization, connecting by the conscience the perceiving subject and the perceived object. However two recent data modify the analysis deeply: - T.G.R. Bower showed that the law of constancy of the size and consequently of the form, was applied by the two month old infant, which suggests a result obtained by the setting concerned of a simple mechanism, asking little training (025). - we could note ourself that the laws of constancy were brought into play in the adult for objects seen for the first time, without any preliminary experiment allowing to allot a size to the object; but that only in the event of binocular vision and not in monocular vision. However, the direct appreciation depth cannot be provided easily by a monocular vision, but is by the telemetric effect of the binocular vision. The law of constancy is thus quite simply explainable by the setting concerned of nonconscious neurological mechanisms and automatically integrating the difference in vision between the two eyes to derive a perception from it from distance. The current neurological studies show the existence of mechanisms of this type. The explanation is then led obviously well beyond the phenomenologic analysis. 3.2.2. Phenomenology and fence. A. Pichot (163) defends the tempting and completely correct idea on the phenomenologic level, that the characteristics of the alive beings are related to an individual lived history, distinct from the history of the environment. The changes are brought back by him to microphone-events selected according to a better aptitude for living. We pointed out to him that such an idea could however be of real interest only at condition of being associated a proposal for an explanation concerning three essential data for which the scientific approach can provide answers: a) the insulation to be it alive of with its environment or is represents a total insulation a control of flow of exchanges which are essential with the life? are b) how the individual microphone-events that are the changes selected if lived is independent of the environment? c) how was the border initially established allowing the first individual indeed living to have lived one independent of the environment? 3.2.3. Phenomenology and Constructivisme. The interest that we carry to the phenomenologic step results us in trying a confrontation between the thesis which we defend and phenomenology, specifying then: - the individual that I am, precede the considered knowledge which I can have of myself, others and surrounding world. The existence does not precede the gasoline, but it precedes the knowledge learned and the contents from conscience. This precession is that of a subject equipped constitutionally with a stable and determining organization biological, capable of an elementary immediate knowledge of ego and world, in all lack of differentiation. - to this knowledge first little by little a knowledge learned from ego is substituted, other and surround which is at the same time a true structuring and a "reflective abstraction" of ego on its own exteriorized or interiorized activity. If learned knowledge presents strong similarities of a subject to another, that is not dependent on the existence of gasolines of universal significance but on a similarity of biological organization. Knowledge thus has a double aspect: a) an immediate knowledge which is of biological and constitutional origin, and which is out of screen between the conscience and external reality. b) a secondary knowledge, learned, conscious which is a true reflexion of our action on the world, but which has the value of the "reduction" postulated by phenomenology, deriving simultaneously from this reduction, the taking into account of our characteristics and of the regularities which characterize the Universe. In conclusion, it seems to us that there is between phenomenology and the epistemological thesis which we will expose a land similarity on the need for a long reflexion for better specifying our true nature and inventing a new image of the world; this reflexion defines learned knowledge. But there is in addition a fundamental difference: - phenomenology postulates a correspondence a priori between gasolines, forms, standards, "gestalts", which at least among first phenomenologists, is present before any reflexion, at the subject of knowledge as in the world. We do not believe in this correspondence a priori. - our point of view is that of the existence of subjective nuns included in our biological organization, that it is necessary for us to release and clarify starting from our action on the world. That makes it possible to carry out inventions or constructions which we then project for better explaining us, and to better explain the world. That marks obviously a joint opening on philosophy and biology. 4.La New Alliance of Philosophy and Science. In 1979, 1. Prigogine and 1. Stengers published a plea (173) in favour of an alliance of philosophy and science. The importance of the concept of "dissipative structure *" isolated by Prigogine and on which we will lengthily return gave them right to the word obviously. It is however quite obvious that several authors had not awaited them to join again a close connection between science and philosophy. 4.1. G Bachelard and History of Sciences. In 1927, Gaston Bachelard opened a fundamental epistemological reflexion, which G Canguilhem was to continue then F Dagognet, Mr. Foucault and Mr. Serres, delivering its letters of nobility being studied of the history of Sciences by drawing two broad outline of criticisms there. - on the one hand, the author finds all the arguments desirable to reject the image positivist of a regular and continuous evolution of Sciences towards a to some extent foreseeable progress (006,007). - in addition, it underlines the defects of the philosophical approaches when they function on themselves, in independence of acquisitions and the evolutions of the scientific thought. Canguilhem expressed the fact a little later: " The philosopher must leave the philosophical cave, if he does not want to condemn to repaître shades, however that the scientists not only see the light but font."(036) The synthesis between these two concerns is centered on a new approach of the truth and reality through philosophy of the inaccurate one. In the center of this philosophy, is the concept of obstacle epistemological. At one moment of the history of Sciences, the analysis shows at which point the reflexion of the scientist is darkened by errors of which it has the greatest evil to get clear: - individual errors which are inherent with the process of operation of knowledge. - errors of the time. The progress of knowledge can be done only in reaction against these errors, and results in successive epistemological ruptures which do not obey a coherent line of unification, simplification or reduction and translate a fundamental discontinuity. These ruptures are synthetic, and cut past. The concern of affirming the cut at Bachelard perhaps comes from the desire to better dispute the continuity of positivism. Piaget sees in the same ruptures, a succession of stages according to an unforeseeable orientation, but where the reflexion on the former stage explains the following stage. For G Bachelard, the error relates mainly to the words which inevitably do not cover the same concept for all the scientists, on the concept itself which is inserted in a context and an intention of the experimentation. There is an overload of contents of the concept which prevents an adaptation to the conditions of use and which give him too much direction, and a variable direction according to users'. Such an approach rejects philosophy by no means. It is on the contrary before the letter, the new alliance of Science and Philosophy. Each one catch separately, the philosophical designs are dogmatic and drying, especially when they neglect the scientific data. On the other hand, a polyphilosophie assistance the dialectical one and correction of the errors. Thus is established an open association towards the future and not a priori, between the current activity scientific, experimental and reflexive, and a historical return on scientific the data processing built before. The benefit is double, melting a philosophy and an epistemology, and bringing a guide in the fields of Science where the risks of skid are particularly important because of the difficulties of the experimental checking. "There are no truths first, it are only errors first."... "the world is my checking, it is made ideas checked in opposition to the spirit which is made ideas essayées."... "L' scientific spirit constitutes like a whole of rectified errors." The checking is a true rational work which answers contradiction: - contradiction between theory and facts - contradiction between various theories. But the approach must be done in the center, "knowledge can change only little by little, under the impulse of a moderate hostility". Ultimately, "knowledge remains always approximate and of the more thorough approximations by slow spiritual modifications the errors in the implicit choices of the former theories reveal unceasingly." Obviously, Bachelard does not deny a positive evolution of knowledge and insistence on discontinuities especially causes to reject the regular growth described by positivism or the orthogenèse finalist which one finds at Theilhard of Chardin. The position of Bachelard seems to us very close to "the orthogenèse noted afterwards", according to the expression of L Brunschvicg. Progress is disengaged from the practice of the existing theories, by a succession of correction of errors and does not obey any preestablished law. "the recurring history, the judged history, the developed history" seem to us synonymous with the formula of Brunschvicg. The point is of importance because this dynamics of progress appears absent analyses of direct or indirect continuators of Bachelard. It is the case at Mr. Foucault or P. Feyerabend. It was also the case of the first approaches of T Kuhn. - for Kuhn (1 15), each time is characterized by a particular design of the scientific theories which reflects sociological conditions and which establishes a model to be followed in any theory. In fact thus sociological factors would be responsible for the scientific revolutions and not the practice of the preceding theories. However, Kuhn appears much affirrnatif in its more recent work. - Feyerabend (064) goes much further, to some extent denying the positive evolution of the theories. There would be according to this author, at any time, a plurality of opposite theories and it would be to the scientist to adopt the theory which appears to him the most adapted to its work. Like pointed out K Popper, it is necessary to take care not to underestimate the risk of a functional autonomy locking up the one time old scientists in a "paradigm", but it is necessary to attach all their importance to the parallel scientific movements which let return of the fresh air and ensure a continued progress. It is in our opinion, to join the theses of Bachelard. On the whole, two concepts emerge from these theses

- the opening towards a progress, but a nondefinable progress a priori, a progress only noted afterwards.

- this progress results in the successive approximations best into better, by elimination of errors. Rather than of elimination of effeurs, we would be remainder personally tempted to speak about a reduction in the indetermination.

Then taking up completely these ideas on our account, we would be tempted to add the concept that progress represents the fixing of a number growing of implicit variables. We will try to show that this total dynamics of progress is explained by the characteristics of cerebral operation, in particular by the narrowness of the field of conscience with respect to the total mental contents.

Today, sixty away years, the initial theses of G Bachelard remain of topicality and cause rich person reflexions. We will find during this work, of very many agreements and recent acquisitions which explain and confirm largely a number of the conclusions of G Bachelard. On the other hand, it seems to us that these conclusions are partially weakened by omissions in two fields:

- G Bachelard does not establish a parallel between the evolution of Science through successive generations and the development of the cognitive thought in the young child, whereas the similarity of the mechanisms appears manifest to us.

- G Bachelard does not include in its analyses the knowledge obtained at the time, in biology and on human neurological operation; it is quite obvious that the multiple discoveries carried out for one half-century have still accentuated this insufficiency.

The constructivism piagétien on the other hand, answers at least the first of these insufficiencies.

4.2. Karl Popper and conjectural knowledge.

K Popper is a contemporary of G Bachelard and it expresses completely comparable epistemological positions (168,169). However Popper is opposed not only to logical empiricism but as much with the ideologies.

- to logical empiricism, it opposes the fact that it is impossible to check "all the" occurrences of an assertion and this fact no assertion of experiment cannot establish a certainty.

- with the ideologies as the psychoanalysis and especially the Marxism, it reproaches a "immunization" which enables them to turn over the contradictory facts, while making them compatible by an ideological adjustment makes a priori.

So it is impossible to reach the certainty and it is necessary to be satisfied to seek the probable one. In addition, it is useless to think that the cognitive steps are done without preliminary theoretical considerations. In other words the assumption is at the beginning and the application of the assumption allows an invalidation or a confirmation; the checking of Bachelard is thus found. For K Popper, "the good" theory is that which defines the experimental results which could invalidate it since the accumulation of different favorable results has only one limit value and is much more difficult to carry out.

But the epistemological theses poperriennes go well beyond the principle of refutability.

- Popper strongly affirms its adhesion with the realism of the objects but it is by no means certain that this realism goes beyond what we described like minimal realism. Popper makes base of realism, the "statement crossed" between several distinct cognitive approaches on the same object, which we will call multicruciality. These cross statements translate steps of observer and implicitly, Popper admits that later experiments can modify the reality division or lead to reinterpretations on the properties granted to an object.

_ Popper affirms a "evolutionary epistemology" of which it finds the roots at J.M. Baldwin, Lloyd Morgan and H.S. Jennings. One can really wonder how Popper made not to meet Claparède or Piaget. Perhaps because of a nativism which made him reject a constructivism calling upon innate knowledge little. "Any acquired knowledge, any training consists in the modification (possibly the rejection) of a certain form of knowledge, or provision, which was as a preliminary already there; and, in last authority, of innate provisions." and Popper still says: "Any development of knowledge consists in the improvement of an existing knowledge which one changes in the hope to more approach the truth."

- Popper affirms the existence of three worlds. World 1 is that of physical reality. World 2 is that of the subject knowing with "its" knowledge, knowledge thus subjective. World 3 is that of objective knowledge. It is a world of ideas, theories, forms, but is not to in no case a world of Platonic realism. It is a world of knowledge which was detached their creators, as the fabric is detached from the spider which secrétée it. "the third world * of Plato was divine, known as Popper; it was immutable and, obviously, truth. There is thus an immense gap between its third world and mine; my third world is a product of the human activity and it is changing. It does not contain only true theories, but also of the false theories, and especially of the opened problems, conjectures and refutations." This world 3 is at the disposal of any knowing subject and it can then cause evolution of this subject as well as subjective knowledge lasting the ontogenetic development. This monde*3 thus corresponds well so that we described above of the concepts, ideas which seem to have a clean existence because they were detached from their creator and the conditions of their birth.

This world 3 comprises consequences which are clean for him and which did not appear at the time of the creation of knowledge. Thus, the man creates the concept of a number and must then discover with properties of the number which it had not considered initially. We create the number then "we discover the prime numbers, and emerges as a consequence the problem of Euclide (the continuation of the prime numbers is it infinite?)"

In addition, and for the same reasons, this world 3 contains necessarily errors, introduced at the time of creation and which must be then corrected.

These epistemological attitudes of Popper appear enthralling to us and let us adhere we essentially to it. The "innate provisions" of Popper are even probably closer to our theses than the "exercises of reflexes" of Piaget. Conversely, can one say that Popper is a true constructivist? We it do not think insofar as Popper well more postulates the evolution of knowledge that of the knowing subject. We should however recognize that Popper is well close to the creed constructivism when there says "It is not nothing direct nor immediate in our experiment, it is necessary for us to learn that we have one me which lasts in time." Perhaps it missed only in Popper taking the dimension of the ontogenetic cognitive development whose it conceived the existence.

4.3. Constructivisme piagétien.

In its principal public works of 1895 to 191 1, J.M. Baldwin (009) defended the principle according to which the individual development and the evolution of the species by natural selection require an active intervention of the organization and cannot be explained exclusively by external forces. So Baldwin substitutes for the function reflex, the "reaction circular" during which the effects of the action on the environment are appreciated by the organization, leading to a adaptatrice activity in the acquisition of the practices or the evolution of the constitutions. They is sixty years before the publications of P. Wintrebert (226,227), the assertion that living it is of its evolution.

Well before the publication in 1933 of the work which was to crown its activities of research, "the genesis of the assumption, experimental study", E Claparède defended the idea that the intelligence represented an adaptation to the new circumstances, substitute with the insufficiency for the adaptations innate or acquired before. This new adaptation is obtained during the action, by gropings which are not tests randomly but specific modifications of the conduits having succeeded previously in circumstances that the subject judges close to those of the action in progress.

Jean Piaget, raises of Claparède, very largely carried out the synthesis of these two notions by exceeding them of the remainder. Biologist of formation but discovering with passion philosophy at twenty years, charged during ten years with a teaching of history of the scientific thought and... father of family, Piaget had any latitude to confront scientific philosophy, psychology and data, which is a fundamental characteristic of its work. The greatest merit of Piaget is to have introduced a historical dimension into any approach of knowledge. In a comparison with other epistemological positions, the following points appear essential

4.3. 1. The ontogenetic development.

The child with the birth is already a "subject" made up. He operates on a purely neurological level, without innate mental contents. But the child is a creator of a psychological development per exercise of the neurological plan. The child with the birth, has a certain number of "reflexes" allowing him a relationship to the environment. The exercise of these reflexes leads to their differentiation, controlled by the organization and allowing the first learned adaptations and the first mental contents.

Learned knowledge is the fact of accumulation of concrete data concerning the answers to the application of the differentiated reflexes, and especially the fact of the acquisition of designs of action derived from the reflexes. These designs are gathered in an increasingly elaborate way, according to stages marked by an increasing mental structuring. A similar biological constitution, the similarity of the physical and social environments only explain how these stages can be identical of one child to the other because each stage does not find its ex...