Bain says that contiguity and similarity are fundamental bases of the memory habit and that they acquire powers in general. He says that writers on mental science have described the law of contiguity by various names. Hamilton terms it the law of ‘reintegration’. “We might also name it the law of association proper, or adhesion, mental adhesiveness or acquisition”. Bain says that the second fundamental property of the intellect may be termed consciousness of agreement or of similarity. It is a great power of mental reproduction, or a means of recovering past mental states. It was noticed by Aristotle as one of the links in the succession of our thoughts.

          But the external conditions alone are not sufficient. Mere observation of similarity cannot give rise to recollection. It is not a sufficient condition, although it is a necessary factor for recollection. The ‘internal competency’ is also necessary.          This refers to the mental preparedness, or, ‘the cognitive urge’. In this sense, Hemacandra says that, though a disposition may have continued for a certain length of time, it does not operate as a cause of memory unless it is aroused.’ In this respect, we may mention McDougall’s emphasis on the cognitive drive as a condition of memory. McDougall says, ‘like all thinking, remembering is a conative activity. We remember and recollect effectively in proportion as we have strong motives in doing so. This truth is too often ignored; we are apt to regard our memory fantastically as a mysterious automatic machine over which we have no control.’ It is notorious that we remember emotionally exciting events better than others; which means that the strength of conation, our interest during any experience, is a main condition of our remembering. There can be no doubt that an explicit volition, purpose, or intention to remember greatly favours remembering and recollecting.

          But even this internal preparedness in the form of interest or conative drive is not sufficient unless some psychic impediments are removed. The fact that our striving to recollect often fails and we get only partial recollection, that we sometimes forget partially or totally, shows that some psychic impediments counteract and come in the way of proper recollection. This is made evident by the study of mental pathology. McDougall says that conation can determine not only memory but also forgetting. Just as desire for an object leads us frequently to remember that object, so aversion to an object (rooted in fear, disgust and painful experiences connected with it) may prevent the remembering of it. It may even make it impossible to recollect it by the most genuine voluntary efforts. McDougall states that thousands of cases of amnesia of this type occurred among soldiers who suffered the horrors of the front during the First World War. Freud also attributes failure to recollect to wishes repressed in the unconscious. In his Psychopathology of Everyday life, he cites instances of forgetting in everyday life. Thus, in order to get effective recollection, it is necessary to remove psychic impediments like aversion to the object, fear and other painful experiences associated with it.

          Such a removal of psychic impediments was, in a sense, mentioned in terms of the removal of karrna. Hemacandra says that, in order to arouse stimulation’s, subsidence and destruction of the obstructive veil of jnanavaraWya karma would be a necessary condition of recollection in addition to observation of similar objects and the conative drive. However, the Jainas mentioned the condition of the removal of psychic impediments in terms of the metaphysical concept of karma and the operation of karma. In fact, the Jainas say that destruction and subsidence of the knowledge-obscuring karma, jnanavaraniya karrna, is a necessary condition of all cognition.

          According to the Nyaya system, while memory has some general conditions, like the original past presentation (purvanubhava), and its mental trace, (sarriskara), it has a number of specific causes which are responsible for retention of the impressions and their recall in casciousness on future occasions. Several factors, like attention (pranidhana), association (nibandha), repetition (abhyasa), and papa and prrnya, are operative as conditions in producing recollection as also in retaining an experience. Chatterjee, in his Nyaya Theory of Knowledge, mentions twenty-three such causes as given by the Naiyayikas.

          The Jainas say that recollection is a valid form of cognition. In fact, it is a source of knowledge, a pramana, because it is never found to be discrepant with fact as in the case of successful activity like search for a thing deposited by oneself. The Vaisesikas and the Advaita Vedantins also accept recollection as valid cognition. Sometimes, an objection has been raised to the effect that recollection is not a source of knowledge, a prarnuya, because it does not cognize the present datum and so has no objective basis. The Nyaya system does not admit memory as a separate source of knowledge, because it is only a reproduction, of past experience in the same form in which it was once experienced. The Naiyayikas say that it is not a preventative knowledge (anubhava). It is only the representation of what was once presented. The object as remembered is different from the object as presented, since the object as presented before, has ceased to exist. The Mimanhsakas also do not regard recollection (smrti), as a pramana, since it gives us knowledge of things only previously experienced; it does not give any new knowledge, but only a revival of the same old knowledge. The validity of remembered knowledge depends on the validity of the previously experienced knowledge.

          But the Jainas say that, while memory is conditioned by the revival of impressions of past experiences, its essence ties in the knowledge of something as ‘that’ in the past (tadityakara). It is the knowledge of what was previously experienced as past. Memory is, in the language of L. T. Hobhouse, assertion of the past as past. That memory refers to a previously experienced object, or that it is an assertion of the past, is known by memory itself. The Jainas say that knowledge of the past given by recollection is valid, like perception, because it leads to successful activity. They also give the criterion for establishing the validity of recollection. If recollection were not valid, inference based on vyapti, the universal relation between the major term and the middle term, would become invalid. Hemacandra points out that recollection refers to an object that has once been experienced, and the reality of the object and not its actually felt presence is the condition of validity for a cognition. If it is contended that the object must be felt as present, as in perception, in order to get valid cognition, we may equally say that perception is also invalid as it is found to lack the criterion of referring to a fact that has been experienced in the past. If revelation of the relevant object be a criterion of validity, it is found to be equally present in the case of memory.

          Again, it has been objected that it would be difficult to understand how an object which is deficient can be a generating condition of recollection. But the Jainas say that this objection is also not valid. Cognition reveals its object when it is brought into being by the requisite condition of the operation of the sense organs and mind and the destruction and subsidence of the knowledge-obscuring karmas, just as light which comes into being on the operation of its own conditions reveals objects, like the jar, though not generated by those conditions. Similarly, if recollection is said to be invalid, one must give up inference also, since inference is not possible without recollection of the necessary concomitance.

          Some Vaisesika writers also contend that smrti (recollection), is a valid source of knowledge. They recognize both smrti and preventative cognition (smrtyanubhava), as a form of valid knowledge. Smrti arises out of impressions of past experience, and it is the knowledge of the individual object ‘as that’, as something previously experienced, like ‘that bathing hat’ and ‘that city of Banners’.

          Vallabhacarya also maintains that smrti is a separate pramana, because it gives true knowledge of certain facts. Although it depends on previous experiences, it cannot be said to be sincerely a repetition of some previous experience. It is something more. It gives the experience of the past experience as past. Awareness of its being past is not a part of previous experience; and memory gives us the knowledge of this new element.

          Among the Western philosophers, Russell, Hobllouse and others recognize memory as a primary source of knowledge. Memory gives us direct knowledge of the past. Russell says that immediate knowledge by memory is the source of all other knowledge concerning the past; without it, there would be no knowledge of the past by inference, since we should never know that there was anything past to be inferred. He says that memory resembles perception in point of immediacy and differs from it mainly in its being referred to the past. Hobhouse shows that memory is neither retention of past experience nor a mere image of past experience, but an assertion of it as past on the basis of such retention and images. Eking also thinks that the view of memory as a direct experience is clearly true if we have any knowledge of the past at all. If we know the past, it is the past we know and not the present ideas of the past. It is a mistake to suppose, as the Naiydyikas did, that we are directly aware of the past, that the past must be, so to speak, bodily present to our mind or occupy the same position as present objects of perception.

          Thus smrtti, or recollection, is considered by the Jainas as valid cognition and a separate source of knowledge. In fact, even inference involves memory, because it cannot take place without the recollection of the universal relation between the major term and the middle term.

          The validity of recollection as cognition is an epistemological problem, although it has a psychological significance. Recall is a revival of past experience. It has past experience as its basis. But we must remember that perception is one kind of mental event, while recall is a different kind of mental event. It is cognitive in nature and an independent source, of knowledge. Drever says that a percept is an event and memory of it a new event. The Jairla analysis of recollection is mainly epistemological, although it expresses the psychological factors involved in the fact of recollection. The Jainas were primarily concerned with the analysis of recollection as a prarnarta. The psychological factors involved in recollection were only incidentally referred to. In fact, all Indian thought gives mainly a metaphysical and epistemological analysis of the problems of knowledge, although psychological factors are incidentally mentioned.


Recognition (Pratyabhijna)

          Recollection (smrti), does not give us a complete picture of memory unless recognition as a factor operates. Complete memory involves retention, recall and recognition.         We may, however, say that retention is a condition of memory, and recall and recognition are not so much conditions of memory. They are forms of expressing the cognition experienced in the past. Remembering may take different forms. The effects of past learning may manifest themselves through the activities of recall or of recognition and they manifest themselves by making it easier to relearn the original experience. Corresponding to these forms of remembering there are different procedures in which memory may be employed as a test for the continued retention of the effects of learning. These are the methods of recall, recognition and relearning. We are not concerned with relearning because it is not a valid source of knowledne as such.

          Recognition was defined as the remembering of something that was presented to the senses. For instance, as Woolworth mentions, we recognize a friend by his visible appearance or by the sound of his voice. His dog may recognize him by the sense of smell. The other senses may sometimes provide cues for recognizing an object already experienced in the past. ‘Cues or signs are used in recognition as they are used in perception. In fact, recognition is a kind of perception.’ McDougall makes a distinction between implicit and explicit recognition. The former is primitive and the latter develops out of it. The dog that runs away at the sight of a man who threw a stone at it, is showing only implicit recognition. The dog does not think ‘this is the man who threw the stone’. For us, the utterance of the proper name of the object is an important part of’ recognition. The similarity of the effect on us is an essential ground of recognition. “The capacity for recognition, and so of all remembering, is at bottom of the fundamental function which James calls ‘conception’ and which perhaps is better called ‘knowing”’.

          The question regarding the nature of recognition was discussed by the Nyaya thinkers. Chatterjee mentions a distinction in the meaning of recognition. It can be understood in two senses. In a wide sense, recognition means understanding the nature of a thing. In this sense it is an ordinary mode of perception. It may be referred to as savikalpa pratyksa. In a narrower sense, recognition means knowing a thing as that which was known before. Pratyabhijfia is recognition in this sense. According to Naiyayikas pratyabhiji is conscious reference of past and present cognition of the same object. I see a jar and I recognize it as something perceived before. Thus I say, “this is the same jar that I saw.”

          It has been maintained by some that recognition is a confusion of two cognition’s, perception and recollection. The Buddhists think that recognition is a mechanical compound of preventative and representative mental states. It is not a single psychosis because it cannot be perceptual in the absence of a sense object contact. Similarly, they say it cannot be a saaska’ra, for there is a sense of thinness in the state of recognition. The Naiyayikas contend that it is a kind of qualified perception giving us knowledge of the present object as qualified by the experience of the past. We see an object and we recognize it as having been seen on a previous occasion. The Mimamsakas and the Vedanta’s support this view. But the Jainas argue that the state of recognition is a simple psychosis. It is synthetic in nature and it is different from perception and recollection.

          The Jainas give prominence to recognition as an important form of cognition. Hemacandra describes recognition (pratyabhijna), as synthetic judgment born of perception and recollection. Perceptual experience and recollection work together to produce recognition. They are both combined to form a synthetic judgment born of perception and recollection. They are, therefore, conditions of recognition. Recognition as a synthetic judgment is expressed as ‘this is that jar’, and ‘this is that cloth’. These are cases of identity. We also get recognition as synthetic judgment which expresses similarity in the form of judgment, as ‘the cow is like the gavaya.’ In this sense, the Jainas make upamana, a form of recognition, and they do not give upamana the independent status of pramarca. We may also get the synthetic judgment of recognition expressed in the judgment of difference. We recognize that the buffalo is different from the cow. Thus, recognition is a concrete psychosis. It is synthetic in nature, expressed in synthetic judgment, like the judgment of identity, the judgment of similarity, and judgment of difference. Perception is the direct and immediate cognition of the object when the object is present to the senses. Recollection is the reproduction, ‘ideal revival’, of what was experienced in the past. It is the emergence of the mental trace to the level of consciousness. When perception and recollection are combined in a particular form to produce synthetic experience expressed in a synthetic judgment, we get recognition. When we get a description like, ‘know him to be Caitra who is shaggy all over the body, who has protruding teeth, who is dwarfish and who has broad eyes and a snub nose’, we make out Caitra when we see him next. Similarly, a man from the North happens to describe a camel as ‘a cursed animal with long crooked neck and with ugly limbs, addicted to feeding on hard sharp bramble’. A man from the south who heard this description happens to see a thing of such description, he then recognizes the animal as a camel in the form of a synthetic judgment, ‘the object in front is a came. The Jainas have emphasized the synthetic nature of recognition as an act of cognition. However, it is a concrete psychosis in which the present and the past, perception and recollection are synthesized. In this sense, recognition is different from recollection, although recognition involves recollection as a factor. In. recognition, the object is present before us; in recollection, what is recollected is not present to our senses.

          A psychological analysis of recognition shows that recognition is a fusion of a percept with an image. Recognition accepts or rejects the object recalled in memory. We cognize when we react to present experience as familiar. The sight of a face, the sound of a note, the smell of a rose, all these may be experienced as being familiar. But we recall a word by speaking it, or we recall past activities after an interval. Hunter makes a distinction between recall and recollection. Recollection involves personal aspects in the memory. Recognition has been described as a mental state which may be definite or indefinite. We may get indefinite recognition in which we only get a feeling of familiarity without getting a definite picture of that experience. Recognition will be definite when it refers to the place and time of the experience. In such _recognition we get, as Titchner said, a revival of the cognition of an object once experienced, associated with a group of other ideas and tinged with a feeling of familiarity. Thus, in recognition, the perception of an object and the recall of the percept are synthesized to produce a concrete psychosis of recognition. The Jainas described such a concrete psychosis as recognition, or pratyabhijna. However, Stout says that recognition in its more primitive form does not require discrimination of the universal from the particular, but only a confused or implicit awareness in which the universal is not separately apprehended as a distinct object of thought. In recognition, there is only a rudimentary judgment of recognition inasmuch as the universal nature of the particular is confusedly apprehended. Yet, there is no judgment in which the subject and the predicate are

mutually sundered from each other. We are not here concerned with the problem of apprehending the distinction between the universal and the particular in perceptual judgment. However, it may not be out of place to say that the Jainas have made recognition a non-verbal form of cognition, in which explicit expression of a judgment in the form of a proposition containing subject and predicate is not possible, although recognition is a form of experience in which we are aware of the similarity or difference of the object which was experienced in the past. In this, we are to understand the description of recognition given in the Pramanarnimarhsa as a synthetic judgment, like the judgment of similarity, identity and distinction, although not explicitly expressed in language.

          But the content of recognition and the content of recollection are different, because recollection only cognizes what has been known before and refers to its content as ‘that’. Recognition establishes the identity of the past datum with the present one.

          The validity of recognition and the nature of recognition as separate source of knowledge, a pramana, has been an important problem in Indian thought. It was very often contended by some schools of Indian thought, tike the Buddhists and the Naiyayikas, that recognition is not an independent source of knowledge, a pramana. The Buddhists say that there is nothing like recognition as a separate source of knowledge, as anything different from cognitive acts like recollection, indicated by the word ‘that’, and perception, indicated by ‘this’. The Naiyayikas say that recognition is a kind of qualified perception in which the present object is qualified by the distinct recollection. of our past experience of it. But the Jainas say that such an objection is not valid, because the object that is known by recognition cannot be comprehended by recollection and perception alone. The province of recognition is the substance which stands out as the identity in and through its antecedent and consequent modes. This identity cannot be the content of recollection, which cognizes only what has been experienced before. But we are aware of the identity of the object experienced in the past with that which is presented to our present consciousness. This identity cannot be cognized only by perception, which is limited to the cognition of the present datum.

          The Naiyayikas maintained, as we have seen, that recognition is nothing but a species of perception. The Samkhya theory also brought pratyabhijna under perception. The eternal buddhi under goes modification by virtue of which it becomes connected with the different kind of cognition involved in recognition. Similarly, the Mimamsakas and the Advaita Vedantins also hold that recognition is a kind of perception. Recognition is that kind of perception in which the object is determined by the name by which it is called, as ‘this is Devadatta’; for, according to Advaita Vedantin, pratyabhijna is a perception of the nirvikalpa type since there is in it no predication of anything about the perceived object, but an assertion of its identity amidst changing conditions. Samkara agrees with the Naiyayikas and the Mimarfasakas in holding that recognition is a perceptual cognition produced by the peripheral stimulation and the subconscious impressions co-operating together. Kumarila agrees with the Naiyayikas in regarding recognition as a preventative cognition, since it is present where there is activity of the senses and is absent where there is no activity of the senses. We cannot treat recognition, he says, as non-perceptual only, because it is preceded by an act of recollection. In recognition also there is a contact of the sense organs with the object, and wherever there is such contact there is perceptual cognition. But the Jainas say that such a view cannot be accepted, because the province of perception is limited to what is actually present and given to the senses. Hence, the identity of the past and the present datum cannot lie within the scope of perception.

          It has been urged that a sense organ, with the help of recollection, does give rise to perception of such identity; and recognition is only a species of perception. But Hemacandra says that this is impossible, because a sense organ cannot go beyond the sphere of the present datum. It is also not true to say that the senses will be able to comprehend identity when associated with recollection, just as the organ of vision acquires additional potency when associated with collyrium. The additional efficiency that might be acquired by a sense organ is never found to overstep its proper jurisdiction. Therefore, recognition is not a form of perception. Nor is it mere recollection. It is not even formed by the mere combination of perception and recollection. It is a synthetic judgment which expresses something more than the mere combination. Therefore, recognition is an independent source of knowledge, a pramana. Hemacandra says that it cannot be said to be lacking in validity, since the lack of discrepancy, which is the criterion of validity, is present in it. On the metaphysical plane, if the identity of the self and the like as determined by the evidence of recognition were to lack objective reality, the logical justification of bondage and emancipation as states of the same ethico-religious aspirant- would become impossible. The sense of identity will have a lease of life only if we accept recognition as a valid source of knowledge.

          This is the picture of the validity of recognition as a source of knowledge. It is mainly an epistemological problem, although it has great psychological significance. Recollection (smrti), and recognition (pratyabhijna), have been described as forms of memory. Memory expresses itself, as we have seen, in recollection and recognition. We have also seen that recognition is a synthetic judgment in which the identity of the present datum with that which was experienced in the past is expressed, although it is still a non-verbal form of cognition. As it is a synthesis of recollection and perception, it would be difficult to maintain that it is an independent form of cognition, a concrete psychosis. Recognition is a form expressing memory. It is sometimes described as a factor involved in memory. And memory is ideal revival. It is mainly reproductive in nature and does not involve transformation of what was revived in accordance with the present conditions. In this sense, it is not possible to say that recognition is an independent form of cognition, although it may be called a psychosis which is synthesized by recollection and perception. However, the Jainas maintain that recognition is not a species of perception nor of recollection. This view is also true because recognition is not just perception nor recollection. It is a synthesis. The synthesis gives the additional quality judgment of the identity of the present datum with that which was experienced in the past. It may also express similarity and difference. However, this problem is more epistemological than psychological.

          Thus, recollection (smrti), and recognition (pratyabhijna), have been considered by the Jainas as valid forms of cognition and sources of knowledge. Retention is a condition of recollection as much as it is a condition of perception. The tendency to endure is a prominent factor in retention; and the absence of lapse is itself a tendency to endure. Retention is also a condition of recollection, because the mental trace retained in the mind makes recollection possible when it is aroused and revived. Modern psychologists make retention, recollection and recognition factors involved in memory. We have seen, as Hunter points out, that recollection and recognition are forms of expressing memory, because memory is not a thing containing parts but the mental activity itself, although ‘faculty’ psychologists made compartments of the mind and memory a faculty of the mind. Even Hume says that an impression makes its appearance in two ways: either it retains a considerable degree of vivacity in its new appearance or it loses that vivacity and becomes an idea. The faculty by which we repeat our impressions retaining the original vivacity is called memory. But modern psychologists do not treat memory as a faculty or a thing but as an activity. We may better talk of remembering rather than memory.

          However, remembering may take different forms. It may express itself through the activity of recall or recognition. In this sense, we may think of recall and recognition as separate and valid forms of memory rather than conditions or factors involved in memory. Srnrti and pratyabhijira would then be the two valid cognition’s. However, such an analysis would be more epistemological than psychological.


Inference (Anumana)

          We now come to another source of knowledge (pramana), which is inference (anumana). The Jainas have mentioned uha, inductive reasoning, and sabda, scriptural authority, as separate pramanas. But these two are not relevant to our discussion, because they have a more logical than psychological significance. Inference, or anumana, is generally recognized by all the Indian systems except the Carvaka as a pramana. Inference and reasoning are expressions of thinking as an activity of the human mind. Modern psychologists have begun to take greater interest in the study of the psychology of thinking. Physiological and psychological analysis of the mechanism of thinking have been carried out by psychologists, especially the Behaviorists and the Gestalt psychologists. William James recognizes that thinking of some sort always goes on. But, as Vinacke points out, the fact of thinking presents two sets of phenomena, (i) the psychological process and (ii) the neural process. The early philosophers in the West gave prominence to thinking as a special and differentiating quality of man. Man was called homo sapiens. Aristotle said that man is a rational animal. The highest form of mental life is reasoning, which utilizes material from sense and imagination, but goes beyond them into the realm of pure ideas. Aristotle worked out a logical system of reasoning which is called traditional logic, Early Greek philosophers gave theories about reasoning as about other mental states, from logical systematization based on introspection rather than from empirical evidence in the modern sense.

          A similar attitude was present in early Indian thought. The Indian philosophers were concerned with building a logical structure of reasoning and incidentally with the epistemic conditions of reasoning, rather than the psychological analysis of reasoning. The theory of knowledge and the analysis of the epistemic conditions of reasoning had for them a pragmatic value. For the Jainas, as for many other Indian philosophers, the ultimate aim was moksa. The realization of moksa is possible by right knowledge as also by right intuition and right conduct. It was, therefore, necessary for them to study the conditions and limitations of knowledge. The Jaina emphasis on the logical and epistemological problems of reasoning expresses the spirit of Indian thought. This study has to be restricted to the nature and conditions of inference as a process of thought. The psychological factors will be referred to, as also the psychological significance of the nature and conditions of inference. This has been included in the discussion because reasoning is a source of knowledge and the analysis of empirical experience would not be complete without understanding the nature and conditions of inference as a source of knowledge.

The Jainas have recognized inference (anumana), as a source of knowledge, (pramana). Most of the Indian schools of thought, with the exception of Carvaka, have given prominence to inference as a source of knowledge. The Carvakas are materialists. They contend that perception is the only pramana. As perception cannot establish a universal proposition, nor can tell us anything about the past and future, perception cannot give us knowledge of vyapti, which is the universal relation between the major and the middle term and the basis of inference. Therefore, the Carvakas say that inference is not a valid source of knowledge as it has no sound logical basis.” But the Buddhists have objected to this contention of the Carvakas. The Buddhists say that the Carvaka refutation of inference is itself a process of reasoning. Similarly, it is by inference that the Carvakas came to know that their views were different and that the other sources of knowledge were not valid. Hemacandra also says that the Carvakas have to depend on other sources of knowledge, like inference, for the validity of their contention. Since perception will not be able to cognize things in the past and future, even with regard to specific direct cognition, the Carvakas will not be in a position to determine the validity or invalidity of cognition to the satisfaction of others. Perception is subjective and so will not be able to establish the objective validity of inference. It was seen earlier that in Plato’s Dialogue, Thecrtetus, Socrates examines the doctrine of knowledge through perception and shows that such a doctrine leads to the impossibility of knowledge. In the Pramanamimansa, Hemacandra says that the validity even of perception can be established only on the evidence of its unfailing correspondence with fact. Hence it follows that Carvaka must have recourse to a different source of knowledge like inference. The Buddhists have accepted inference as the other source of knowledge. In fact, the Buddhists make all non-perceptual cognition necessarily of the nature of inference.

          The meaning of inference has been a difficult problem in Indian thought, though there has been general agreement on the essential nature of inference. The Jainas say that inference is mediate know ledge. It is knowledge obtained through some other knowledge. Hemacandra says that inference is the knowledge of the major term on the strength of the knowledge of the middle term. The Jainas hold that anumana is the process of knowing an unperceived object through the perception of a sign and the recollection of its invariable concomitance with that object. It is called anumana because it is the organ of subsequent (anu) cognition (mana). The knowledge of the major term which is of the nature of authentic cognition of a real fact and which arises from a middle term either observed or expressly stated, is called inference. It is really cognition which takes place subsequent to the apprehension of the middle term and the recollection of the necessary relation of the major term and the middle term. In the Jaina Tarkabhasa, a definition of inference as given in the Pramanamimansa is mentioned. The Nyaya system has worked out an elaborate system of inference. It is primarily a study of inference. Vatsyayana, in his exposition of the process of reasoning described by Gautama, asserts that the process of reasoning is extremely subtle, hard to understand and only to be understood by one of much learning and ability. Keith says that the admission of such a nature is important, because it shows how difficult were the first steps of understanding the process of reasoning. Anumana, literally, means knowledge which follows from some other knowledge. It is knowledge of an object due to the previous knowledge of some sign, linga. The previous knowledge is the knowledge of the sign which shows the universal relation between the major and the middle term. Anumdrra has been defined by the Naiyayikas as knowledge of an object not by direct perception but by means of the knowledge of a liriga, or sign, which expresses the relation between the major and the middle term. Bhasarvajna defines inference as a means of knowing a thing beyond the range of senses through its inseparable connection with another thing which lies within the range of senses. Gangesa defines inference as knowledge which is produced by some other knowledge. The object of inference is the knowledge of some fact which follows from the knowledge of some other fact. By means of anarmarra we want to know that which may not be perceived but which is indicated by previous perception. For instance, anumana leads to the knowledge of a hill having on the basis of the perception of the smoke on the hill.’

          All systems of Indian thought, except the Carvaka, believe that inference is a process of arriving at truth not by direct observation but by means of knowledge of the vydpti, the universal relation between two things. The Buddhists believe that inference consists in perception of that which is known to be universally connected with another thing. Such a connection is either due to the principle of causality or to the principle of identity. According to the Vaisesikas, inference is knowledge derived from the perception of a liriga, or sign, which is uniformly connected with something else, such as cause, effect, co-effect and correlative term. The Sdmkhya, Yoga, Mimamsa and Vedanta systems define anumana as knowledge of one term of a relation which is not perceived through the knowledge of the term, but which is explicitly understood as related to the first term. In this sense, inference is a process of thought in which from something known we arrive at something unknown.

          In Western thought, Miss Stebbing distinguishes inference from suggestion and recollection. However, it is difficult to distinguish precisely between those experiences in which inference is not involved and those in which it is. Psychologists do not agree as to where the line should be drawn. It is not, however, legitimate to distinguish, she says, between two kinds of inference as psychological inference and logical inference. All inference is psychological, for inference is a mental process; but its validity depends on conditions that are logical. Inference, then, may be defined as a mental process in which a thinker passes from the apprehension of something given- datum-- to the apprehension of something related to the datum in a certain way. The datum may be a sense datum, a complex perceptual situation, or a proposition. The datum of an inference can always be expressed in a proposition. Hence, inference may be said to be a mental process in which a thinker passes from one or more propositions to some other propositions connected with the former in a certain way. Western philosophers and physchologists are not agreed as to the essential marks of reasoning. On the one hand, there are philosophers who regard reason as quasi-divine and a spiritual function, while the materialists and some modern philosophers like Strong, Santayana and Russell have thought of reasoning as merely a complex process of associative reproduction essentially determined by the physicochemical process in the brain proceeding according to the purely mechanistic laws of habit. From the point of view of psychology, McDougall says that the essence of all reasoning is that a judgment and a new belief are determined by beliefs already established in the mind. If the old beliefs are true and the reasoning process correct, the new belief is true and becomes an effective guide to action. In this he includes inductive reasoning also. In the most striking cases, the new belief is derived from a complex chain of processes front a previously established belief : as when the astronomer Adams arrived at the belief that a hitherto unseen planet would be seen at a certain position in the heavens if a sufficiently powerful telescope were directed to that spot. Some modern psychologists have tried to reduce the whole thinking process to neural activity. They have made it implicit talking. But this problem is not relevant to our purpose.

          Inference has been distinguished from perception. It cannot be identified with perception, although both are equally valid sources of empirical knowledge. Perception is independent of any previous knowledge, while inference depends on previous perception. It is sometimes defined by the Naiyayikas as knowledge which is preceded by perception. It is based on the perception of the relation between the middle and the major term as subsisting in the minor term. Secondly, perception is due to the contact of the sense organs with an object.          Hence, perception is limited to the cognition of the present. But in inference it is possible to get knowledge of the past and future in addition to the knowledge of the present. Perception, therefore, is direct immediate knowledge, while inference is mediate knowledge. Hemacandra says that perceptual cognition arises out of the datum present to the senses. It is incapable of taking cognizance of what has preceded and what is to follow. Therefore, it cannot discern a characteristic capable of determining the validity or invalidity of the individual cognition’s occurring before and after. Similarly, it is not possible by means of perception to have acquaintance with what passes in other people’s minds. Udyotakara mentions this point when he makes a distinction between perception and inference. Perception is confined to objects of the present time and within the reach of the senses, while inference relates to past, present and future. Perception and experimental observation do involve an element of inference in that the perceived element is interpreted. Samkara says that where perception is available inference has no place. Buddhists made another distinction between perception and inference. For them, perception gives, though inexpressible in words, the peculiar character, (svalaksana) of the momentary object, while inference deals with ideal generality (samanya laksana). But the Naiyayikas do not accept this distinction. For the Naiyayikas, perception gives us knowledge of the individual in its concrete detail as well as its generality, while in inference we deal with generality only in an abstract form. For instance, we have, on the one hand, before us fire which we perceive; on the other hand, we infer the existence of fire past, present and future as generally connected with smoke.


Types of Inference

          Indian logic does not make a distinction between deductive and inductive inference as separate forms of inference. Rather, an inference is a combined deductive and inductive process. Similarly, the distinction between immediate and mediate inference is also not found. All inferences are in the form of categorical syllogisms; and they have both formal and material validity. A distinction between deductive and inductive inference is psychologically inadequate. Vinacke points out that it has become conventional to recognize two broad areas in logic: formal logic, which is called deduction; and scientific method, which is called induction. ‘It is now commonly recognized, however, that these distinctions break down in the actual process of reasoning, although deductive inference is often the only observable process in formal syllogistic situations. If syllogisms are extended into everyday life so that their origins can be traced, inductive processes occur’. Dewey has endeavored to rid logic of such distinctions as deductive and inductive inference, because both kinds of esquire are fundamental in science and such a distinction is possible through intellectual analysis. Even the division of inference into immediate and mediate is not psychologically sound. The process of inference is always uniform and one. It is the process of thought in which from something which is already known we arrive at something relating to something new which is not present to the senses. In this sense, immediate inference is only a brief expression of the process of inference. The main function of mediate inference is to communicate systematically one’s own reasoning to others with a view to convincing them or rather with a view to creating similar beliefs in others. For the sake of our own knowledge and conviction it is not necessary to establish an elaborate system of reasoning in the form of syllogism. In this sense, the division of inference into immediate and mediate has no psychological significance, although it may have logical importance and validity.

          In this sense also, it may be said that a distinction has been drawn, in Indian thought, between inference as inference for oneself (svartha anumana) and inference for others (parartha anumana). Almost all Indian systems have made such a distinction. In the Pramanamimamsa inference has been similarly distinguished. Inference for others is described as syllogistic in nature. Inference for oneself is subjective and ‘is calculated to remove personal misconception’, while syllogistic inference ‘is capable of removing the misconception’ of another person. Subjective inference is also based on the knowledge of the relation of the major with the middle term. Still, it needs to be expressed in elaborate syllogistic form. The Naiyayikas made three classifications of inference: (i) svartha anumana and parartha anumana; (ii) purvavat, sesavt and samanyato drstam; and (iii) kevalanvayi, kevala vyatireki and anvaya vyatireki. Keith points out that the distinction in inference as svartha and parartha was wholly unknown to Gautama and Kanada but was accepted by the Syncretist School. The classification of inference into svartha and parartha is a psychological classification which has in view the purpose which the inference serves. With reference to the purpose, all inferences are either meant for acquiring some new knowledge for oneself or for the demonstration of a known truth to others. In the svartha inference, a man seeks to reach a conclusion for himself. In parartha inference, the aim is to demonstrate the truth of the conclusion to others. The conclusion is justified with the help of the middle term. For instance, in the parartha anumama a man, having inferred the existence of fire on a hill, lays it down as a thesis and proves it for others. The other two classifications mentioned by the Naiyayikas have rather logical significance than psychological value. Regarding the distinction between the svartha and parartha anumana, it may be pointed out that inference for oneself is notional (jnanatmaka), as Dharmottara stated. Inference for others is verbal (S’abddtmaka). Keith points out that the Nyaya view of the distinction shows that, in inference as communicated by the syllogism, that is parartha inference, the hearer must perform the necessary mental operation which the teacher has already preformed and which he now helps by syllogistic exposition the hearer to perform for himself. Therefore, it can be said that the svartha inference deals with the process of inference and the parartha inference is the formal expression in syllogistic form. The first is characterized as artha_ rupatva, as Sivaditya showed, the other as sabdarupatva.

          Vinacke points out that, if deduction is regarded as a method by which already existing generalizations are used, it is found that deductive situations are widely encountered in everyday life. They are not always evident as such. They often occur in a disguised and incomplete form. He says that, in general, two aspects of the problem may be distinguished. On the one hand, there are conditions under which the individual argues with other people; on the other, there are more or less public arguments to which the individual is exposed. In the first situation, we are obliged to make assertions, develop arguments and state conclusions with a view to communicating and demonstrating them to others. In the second type of situations, we find ourselves reading in the newspapers or magazines arguments presented implicitly or explicitly in deductive form. In all such situations the rules of logic are valuable grounds for valid arguments. Although this distinction between the two deductive situations presented by Vinacke does not exactly correspond to the svartha and parartha anumana, the analysis of the first situation corresponds to Parartha anumana. Parartha anumana expresses itself in elaborate argument in syllogistic form.

Conditions of Inference

          The aim of inference is to attain some new knowledge of a thing on the basis of whatever has been already known. It arises out of the necessity to know something more, as also out of doubt and anxiety regarding the thing to be known. Where perception is available, inference is not necessary, because we need not reflect much to know objects present to our senses. Inference is not possible regarding either things unknown or things definitely known. It functions only with regard to things that are doubtful. Doubt is a condition of inference. It implies not only absence of certain knowledge about something, but also a positive desire or will to know it. Modern Naiyayikas do not accept this view, because, they say, there may be inference even when there is no doubt and in the presence of certainty. Similarly, there may be inference even when there is no will to infer. The inference aims at proving that which is yet unproved, as there is a desire to prove the object. At the same time, as Hemacandra says, it is incapable of being contradictory. Therefore, it is generally accepted by all schools that a logical discourse does not come into play in regard to matters which are unknown or definitely established. That a state of doubt is a motive of inference is very often recognized in psychology and philosophy. Doubt sets us thinking and gives rise to efforts towards the solution of a problem. The Jaina philosophers, in fact all Indian philosophers, have stated that desire to know is an additional factor for inference. So, too, Miss Stebbing shows that doubt is a psychological condition of inference.

          Inference consists in establishing the relation between the major and the minor term. Knowledge of such a relation depends on the knowledge of the vyapti, universal relation between the major and the middle term. Knowledge of the major term, which is of the nature of authentic cognition of a real fact and which arises out of the middle term either observed or expressly stated, is in fact called inference. It is a cognition which takes place subsequent to the apprehension of the middle term (linga grahana) and the recollection of the vyapti. Regarding the vyapti Das Gupta points out that the Jainas, like the Buddhists, prefer antarvyapti (e. g., relation between smoke and fire) to bahirvyapti (relation between the place containing smoke and the place containing fire). The Buddhists showed that vyapti may be based on essential identity, causality, tadatmya and tadutpatti. Experience cannot be the sure ground of vyapti. But the Vedantins make it the result of inductive generalizations based on simple enumeration. The Naiyayikas agree with the Vedantins in showing that vyapti is established on the basis of uncontradicted experience.

          Just as inference depends on the knowledge of the vyapti, it also depends on the knowledge of the relation between the middle and the minor term. This is often called paksa dharmata. In inference, the minor term becomes related to the major through its relation to the middle term. Chatterjee points out that, while the validity of the inference depends on vyapti, the possibility of inference depends on the relation of the minor with the middle term which is also called paksata. Vypti is the logical ground of inference, while paksata is the psychological ground of inference. Kesava Misra explains the process of inference as follows: In the first stage the operation leads to the perception of invariable connection between the major and the middle term. This is arrived at from frequent observations of the occurrence of the two in the past.

          For instance, smoke is observed on a hill. We then remember the relation which perception has established between smoke and fire. This gives rise to reflection in the form that there is on the hill smoke, which is always accompanied by fire. Then we arrive at the inference that there is fire on the hill. Keith points out that this value of the conception of inference as a mental process is enforced in minute detail by the Nyaya school. From another point of view, stress is laid on the fact that the subject, the minor term, must be something regarding which there is a desire to establish something else. This desire may be for one’s own satisfaction or for that of others. Bosanquet also considers such a mental activity of inferring as the decisive feature of inference. 

The conditions of inference have been discussed by modern Western logicians. Russell seems to think that the psychological element of our knowledge of the propositions and their relations, is not a necessary condition of inference. Validity of inference mostly depends on the logical condition of the implication between propositions. We infer one proposition from another in virtue of a relation between two propositions ‘whether we perceive it or not’. The mind, in fact, is as purely receptive in inference as commonsense supposes it to be in perception. of sensible objects. But W. E. Johnson and Miss Stebbing have recognized both the psychological and the logical conditions of inference. The logical conditions consist in the relation between the propositions. They are called ‘ the constitutive conditions’. The psychological conditions have been called ‘the epistemic conditions’ of inference. They refer to the relation of the propositions to what the thinker may happen to know. Earlier in the chapter, Johnson says that inference is a mental process which, as such, has to be contrasted with implication. The connection between the mental act of inference and the relation of implication is analogous to that between assertion and proposition. Miss Stebbing also shows that inference involves both the constitutive and the epistemic conditions. The epistemic conditions relate to what the thinker who is inferring knows.

The question regarding the special cause of inference (karana) that brings about the conclusion in inference, has been discussed by Indian logicians. According to the Buddhists, the Jainas and some Naiyayikas, it is the knowledge of the liriga, the middle term, that leads to the conclusion. The middle term known as such is to be taken as the karaha or operative cause of inference. R. S. Woolworth says that reasoning very often depends on the use of the middle term. The Mimansakas and the Vedantins believe that the knowledge of vyapti is a cause of inference. According to them, the knowledge of the universal relation between the major and the minor term is received in our mind when we see the liriga or the middle term as related to the paksa or the minor term.