On Experience Occluding Structures
1.1 How did the proponents of the ‘flat-earth’ hypothesis react to the claim that the earth was spherical? By suggesting that, if the earth were to be round, at some point in a sea or land travel one would have to fall off the surface of the earth. Because such an incident never occurred, they reasoned, it was preposterous to entertain the idea that earth was spherical.
1.2 When Galileo argued the helio-centric theory in his Dialogues on the Two World Systems, his opponents refuted it by putting across the following consideration: if earth revolved round its axis, how does it come about that we do not fly off the surface of the earth? Or feel dizzy? Or why does an object thrown up during a boat travel fall inside the boat and not miss it entirely? Such experiences, they suggested, goes to show that Galileo’s theory was absurd.
1.3 When the motion of the planet Mercury threatened to become an anomaly to Newton’s theory, the scientists did not think that Newton’s theory got refuted by the observations. Instead, they postulated an ad hoc hypothesis: there must be another planet (Pluto, if my memory serves me right) in the vicinity causing Mercury to deviate from its predicted path. They could not believe that Newton’s theory, which so elegantly accounted not only for planetary motions but also the ebb and flow of the tides, was wrong.
1.4 These are but three of many such stories told by philosophers of science, while discussing the adequacy of theories of scientific growth. What is of interest to me in such anecdotes is the nature and role of ‘experience’. It plays a very important role in each of the above examples: in the first two, there is an appeal to ‘experience’ in order not to accept accounts which appear ‘contra-experiential’; in the third case, an ad hoc hypothesis is proposed in order to ‘save the experience’. Despite their dissimilarity, there is a cognitive attitude common to all three: threats to experience are rejected. (In philosophy of science, this attitude is called as ‘saving the phenomena’. My reason for not choosing the accepted terminology will become evident later on.)
The constraining notion of moral rule appears to appeal to another, more deeply rooted idea from the Christian religions: man is an immoral creature. Thus, it follows, in the absence of moral rules there would be no end to immoral behaviour. Perhaps, human civilisation itself would be impossible in the absence of prescriptive rules for human actions. (This cluster of ideas is not as explicitly present as I have just formulated it.) I have been puzzled by these claims for some time now, and each time I have had discussion with my students on this, my puzzlement has only increased and not decreased. What is the puzzle?
3.1 Let me recount the structure of the discussions I have with my students. The discussions begin with their belief that moral rules constrain the immoral strivings present in each one of us. They think that they do not steal, murder, or rape the female students because of their moral education. In its absence, they feel, they would have been very immoral creatures. Why do they think so? “Look at all the wars, genocides, murders, thefts, and sundry criminal activities”, they say. Let us grant the assumption. How often, I ask the male students, have they had to struggle against the urge to rape their female student-colleagues? Hardly, if ever. The idea does not even occur to them, leave alone that they struggle with this urge. How many female students have been raped by the male members of the university? They are not aware of any such incidence, which means that, even if it has happened, it is not a frequent occurrence. Surely, if moral rules are constraints on our ‘urges’, more would have lost the struggle against these urges than now. A half-convinced, half-sceptical assent follows. How many times did they have to restrain themselves from stealing in a shop, murdering a fellow human being? Apart from a very occasional ‘desire’ to steal from a super-market, it transpires, they are remarkably free of any awareness of such struggles. However, if moral rules constrain us, surely, we must chafe against them more often than we are aware of doing it? Again a half-convinced, half-sceptical assent follows.
Which is easier? Stealing from a super-market, or paying for the merchandise? Remarkably enough, the first answer is, invariably, ‘stealing’. Why, then, do they not do it often? ‘Our moral upbringing’ or the ‘fear of punishment’. The subsequent step is to draw their attention to the fact that it is not easy, but very difficult to steal: thumping hearts, sweaty palms, dizzy heads, meaningless and repetitive loitering in the shopping mall, continuous fearful looks, etc. are the prerequisites, surely. The whole experience is basically so unpleasant that, even if followed by an exhilaration of success, it leaves one feeling week-kneed for hours after the experience. If this is the case, how could they say that it is easier to steal than pay for the merchandise? It could not be the case that their appeal to ‘moral upbringing’ accounts for the ‘difficulty’ in stealing. As children, most of them have hardly been instructed not to steal, any more than not to kill or rape. Even if they have been, they were also instructed, more often and without a great deal of success, to tidy up their rooms, clean up the toys after playing, etc. Why should a simple instruction ‘do not steal’ make it so difficult for them to steal, when no amount of nagging helps them keep their rooms clean? ‘Well, if you look at it that way …’ Which other way could one look at it? There does not seem to be an answer to this question.
If one were to add up all their ‘immoral acts’ and compare it to the ‘moral’ (or the ‘neutral’), what would their relative weights be? Immoral acts (of the petty sort) seem to form a small percentage. Were we to do the same with respect to human history? ‘Well, if you look at it your way, the number of immoral acts must be pretty insignificant when compared to the number of ‘moral’ or ‘neutral’ acts…’ Is it not the case that we would expect to find more such acts if moral rules constrained us? Yes. How then can one say that human beings are ‘immoral’ and moral rules constrain us?
It is no part of my argument to say that man is basically ‘moral’. But what does puzzle me is this: how could they entertain a belief that is belied almost everyday? May be, they have not ‘reflected’ on this issue before. In that case, what about the philosophers? Why do they maintain a belief that is perhaps the most refuted claim in the world? In this article, I hope to seek at least the fragment of an answer.
3.2 What is of interest in these discussions is their unwillingness to let go of their belief in the nature of moral rules or their constraining role. They desperately seek arguments to refute me and when they fall silent and appear to grant me my claim, it is not because they are convinced but because they have no further arguments to offer. It has rarely happened that a student shares my puzzlement. Consequently, I am even more puzzled: what exactly is going on here? Why the resistance or even the downright hostility that gets expressed occasionally?
4.1 Let us look at these cases not in terms of their (implicit) beliefs, but in terms of how they learn what they have learnt. It is their experience that people do not go around murdering, stealing, and raping each other. This experience gets structured in terms of the explanation about the nature and role of moral rules. ‘Experience’ has the structure it has because of the structure of the explanation. In other words, there is no difference between the (structure of the) cognitive scheme and the structure of experience itself.
4.2 If such is the case, the resistance in each of the above four cases becomes understandable. The theory of flat earth structured the experience of not falling off the earth; geo-centric theory structured the experiences of not flying off, not feeling dizzy, etc.; Newton’s theory structured the experience of the ebb and flow of the tides; the moral theory structured the experience of people going about their shopping, greeting each other in a ‘normal’ way. That is, experience is an absolutely structure-hugging fabric. The structure of the cognitive scheme is the structure (or the figure) in this case. To challenge the structure of the cognitive scheme is to challenge the structure of the experience itself. The structured experience threatens to lose its structure, if the cognitive scheme gets discarded. Unless, of course, an alternative cognitive scheme is present that reduces this disorientation.
4.3 There is a dangerous drift to this argument. It seems to be heading in the direction of the following claim: all structured experiences derive their structure from the structure of cognitive schemes (or theories). Therefore, either no ‘pure experience’ is possible or the way to have it is by discarding all theories. In this recognizable route, there is an inevitable consequence: knowledge becomes hindrance to experience. Scientific theories end up masking our experience, and all theories entail dogmatism on our part. Of course, such ‘mysticism’ is no part of what I want to say.
4.4 Therefore, let me demystify what I have said so far by reformulating it as an ‘error’: people confuse an ‘explanation’ of an experience with the ‘experience’ itself. Consequently, when they meet a criticism of their explanation of a given experience, they mistakenly see it as a criticism of the experience itself. Hence their resistance to criticisms. They are, in other words, ignorant of the distinction between ‘having an experience’ and ‘having an explanation for the experience’. This formulation might appear more perspicuous. Let me, therefore, stick with this for some time.
4.5 Ordinary folk might be oblivious to this distinction; how about the extra-ordinary folk like scientists and philosophers? Why do they commit the same mistake with almost the same frequency? What makes it difficult even for them to keep this distinction in mind? Or, more generally put: where does this error come from? Why does it persist? What is the nature of this particular ignorance (of the distinction) that makes us forget the distinction? Why does not learning about this distinction eliminate it once and forever from the republic of knowledge?
In the Indian traditions, there is another equally dominant notion of ignorance present: it is a positive force of some sort. From the Buddha through the Advaita of Shankara to the Bhakti traditions, one refrain can be heard over and over again: ignorance causes suffering. It is very easy to think that ‘ignorance’ refers to the absence of information as well: one has to merely learn this or that doctrine, and one becomes ‘knowledgeable’ as well. This is how Indian traditions have been looked at in the course of the last few hundred years at least. But this interpretation is not acceptable: how can an absence (of information) cause anything? In fact, one of the widely accepted criteria to speak of ‘existence’ is the ability to act as a cause: only an existing something (object, event, or whatever else) can cause something else. How can ignorance cause anything (let alone some such thing as suffering), if it is merely the absence of information?
At first sight, Indian traditions seem to be saying something analogous. The Sanskrit words literally mean ‘knowledge’ and ‘not-knowledge’. ‘Not-knowledge’ appears to refer to false beliefs: that the agent exists, the agent acts, etc. Often, it is called illusion as well. But the most compelling reason for not accepting this prima facie appearance is the following consideration: without exception, almost all these traditions identify ‘non-knowledge’ as the biggest obstacle to knowledge. If ignorance were to be merely the absence of information, or the absence of knowledge, how could it be an obstacle to gaining knowledge? It is, of course, contingently possible that some kinds of false information prevent you from having knowledge. But the Indian traditions do not speak of these kinds of contingencies. They appear to suggest that inherent in ignorance is its capacity to actively prevent you from attaining knowledge.
8.1 In the first two examples, it is obvious how the protective belt functions: experiential structures are rendered immune to criticisms. Any consideration, whether factual or theoretical, is repulsed from damaging the structured experience. The only hope of change here is the presence of an alternate cognitive scheme that structures the experience differently. In the fourth example, the story is pretty much the same as in the first two cases.
8.2 If we look at the results of the third example, two things occur because of the protective belt. Newton’s theory becomes impervious to the factual observation regarding the perturbation of the motion of Mercury. The protective belt does something that is, at first sight, curious as well: it enables the generation of a new piece of knowledge (at the moment, it is irrelevant whether that piece of knowledge was ad hoc or whatever).
8.3 These illustrations are all I need to formulate my hypothesis. The protective belt of Lakatos is actually a learning strategy. He called it a protective belt; I call it ignorance. Ignorance is a learning strategy which immunises. What does the immunisation consist of? An identification of the structure of the experience with the structure of the explanation (or the cognitive scheme) itself. The protective belt or the learning strategy (that ignorance is), then, immunises by inducing the identification of the structure of the experience with the structure of the explanation itself. It also enables an active repulsion of theoretical and factual criticisms by deflecting the criticisms. That is, the learning strategy relocates the force and foci of such criticisms.
9.1 What we learn as we grow up in our societies and cultures are salient diversities. We not only learn to see the world in terms of salient diversities, but we also learn the attitude of working with them as long as they appear to work well. That is, we are prone to retain the salient diversities we have learnt; it is also not possible to enumerate them. Such an exercise is not even necessary: we need not know every plank in the boat; what we need is the ability to localise the plank which has sprung a leak. If we have this ability, it more than compensates for the lack of explicit knowledge of each and every plank. This ability does not just compensate without being positively useful: it is an enormous waste of cognitive resources to learn each of the component part individually. Furthermore, we do not have the time to do this either. If the boat springs a leak while we are acquiring the knowledge of some or another individual plank, we are well and truly sunk. That is to say, it makes enormous evolutionary sense to have the learning ability to localise and repair the individual planks instead of having to learn each one of them individually.
9.2 The second aspect of this learning ability is often called the ‘pragmatic presupposition’ for action. When we work in the office, we pragmatically presuppose that ‘everything else’ continues to be what it is. When we write on a piece of paper, we ‘pragmatically presuppose’ that there is a causal connection between the pen moving on the paper, and the appearances of marks on the paper. When Hume wrote his ideas down, he ‘pragmatically presupposed’ that moving his hands on the paper this way, and not that way, had specific effects; that the words did not disappear once they got penned; etc. Of course, this is but a name, but does not explain. However, it does tell us that this ‘attitude’ is deeply rooted in each one of us. When we are in some part of the boat, we take it for granted that the rest of the planks continue to be what they were and behave the way they are supposed to. We cannot question all the salient diversities at the same time. Not only because it is cognitively impossible, but also because we do not explicitly know what they are. We take the existence of other salient diversities for granted, assume their ‘correctness’, while we interrogate some salient diversity. This attitude is a precondition for learning: we have to hold the rest of the salient diversities stable, if we want to learn or alter some specific salient diversity. That means to say that the learning ability creates the pre-condition for learning as well.
9.3 To summarise: the learning ability helps us localise the plank that springs a leak; it also creates the pre-condition for being able to do so by developing an attitude of holding the rest of our ‘experience’ steady and stable.
This learning ability, I now want to suggest, is an evolutionary inheritance. Imagine the case of you taking your dog for a walk. Neither the dog nor you ‘think’ that the house you live in disappears the moment you leave it. Yet, you do not ‘assume’ that it will be still there when you come back; neither does the dog. In fact, you do not think about this at all. Both of you ‘take it for granted’. This attitude is what enables both you and the dog to learn. Suppose the house you lived in disappears by the time you are back from taking a walk. In that case, both the dog and you are ‘puzzled’: as a human being, you fish around for probable explanations; the dog simply circles around, smelling the ground for ‘clues’.
9.4 As I see it, ignorance as a learning strategy springs from this evolutionary inheritance. It holds ‘experience’ stable by inducing an identification of explanatory structures (or the structure of the cognitive schemes) with the structure of experience itself. It helps localise the leaking plank by meeting ‘criticisms’ through the generation of hypotheses. The first entails that, at times, the hypothesis becomes ad hoc and that ‘criticisms’ are met by deflecting the force and foci of such criticisms. But this is an inevitable side-effect of any heuristic, any learning strategy. No heuristic is infallible. Yet, this strategy is retained because it originates from an enormously successful learning ability, which creates the pre-condition of learning. This precondition for learning is also generated by the learning strategy that ignorance is: absence of knowledge. (‘Holding experience stable’ is simply another description for ‘absence of knowledge’, as a moment of reflection will make obvious.)
9.5 Without ignorance, we cannot learn. It is our learning strategy that generates ignorance; ignorance is the price we pay for being able to learn. That is, we see how ignorance is both a presupposition and product of the process of learning. Learning does not only produce knowledge (about some specific aspect) but it also reproduces ignorance (in other aspects). They are not unrelated to each other, which is not how they look prima facie. (‘Of course, learning about black holes does not contribute to your understanding your digestive processes’.)
9.6 It is important to note that the evolutionarily inherited learning ability is itself not ignorance. (Human) ignorance is a specific implementation (or ‘translation’) of this learning ability. Ignorance (I speak only of human ignorance) is a learning strategy that implements the evolutionary ability in a particular way. This entails that there could be other ways of implementing ‘or ‘translating’ this enormously successful ability to learn. What other ways could there be? This question has been identified, albeit in a slightly different guise, as the ‘frame problem’ in Artificial intelligence. Daniel Dennett (1984: 129-130) tells a long but beautiful story about this, which goes as follows:
“Once upon a time there was a robot, named R1 by his creators. Its only task was to fend for itself. One day its designers arranged for it to learn that its spare battery, its precious energy supply, was locked in a room with a time bomb set to go off soon. R1 located the room, and the key to the door, and formulated a plan to rescue its battery. There was a wagon in the room, and the battery was on the wagon, and R1 hypothesized that a certain action which it called PULLOUT (WAGON, ROOM) would result in the battery being removed from the room. Straightaway it acted, and did succeed in getting the battery out of the room before the bomb went off. Unfortunately, however, the bomb was also on the wagon. R1 knew that the bomb was on the wagon in the room, but did not realize that pulling the wagon would bring the bomb out along with the battery. Poor R1 had missed that obvious implication of its planned act.
Back to the drawing board. ‘The solution is obvious’, said the designers. ‘Our next robot must be made to recognize not just the intended implications of its acts, but also the implications about their side effects, by deducing the implications from the descriptions it uses in formulating its plans.’ They called their next model, the robot-deducer, R1D1. They placed R1D1 in much the same predicament that R1 had succumbed to, and as it too hit upon the idea of PULLOUT (WAGON, ROOM) it began, as designed, to consider the implications of such a course of action. It had just finished deducing that pulling the wagon out of the room would not change the colour of the room’s walls, and was embarking on a proof of the further implications that pulling the wagon out would cause its wheels to turn more revolutions than there were wheels on the wagon – when the bomb exploded.
Back to the drawing board. ‘We must teach it the difference between relevant implications and irrelevant ones’, said the designers, ‘and teach it to ignore the irrelevant ones’. So they developed a method of tagging implications as either relevant or irrelevant to the project at hand, and installed the method in the next model, the robot-relevant-deducer, or R2D1 for short. When they subjected R2D1 to the test that had so unequivocally selected its ancestors for extinction, they were surprised to see it sitting, Hamlet like, outside the room containing the ticking bomb, the native hue of its resolution sicklied o’er with the pale cast of thought, as Shakespeare (and more recently Fodor) has aptly put it. ‘Do something!’ they yelled at it. ‘I am’, it retorted. ‘I am busily ignoring some thousands of implications I have determined to be irrelevant. Just as soon as I find an irrelevant implication, I put it on the list of those I must ignore and …’ the bomb went off.
All these robots suffer from the frame problem.” (“Cognitive Wheels: the Frame Problem of AI.” In Christopher Hookway, ed., Minds, Machines and Evolution, Cambridge: Cambridge University Press, 1984, pp. 129-151.)
9.7 It must be clear what R2D1’s problem is. It must first actively consider (or focus upon) an implication before ignoring (or neglecting) it. Human ignorance does both simultaneously: while focussing on some specific thing, it neglects everything else. These two facets of ignorance, directed towards two different things, are difficult to implement in a system, where everything requires to be represented explicitly. In my frame work, the question of AI becomes: how do you explicitly represent ignorance, when it consists precisely of not being explicitly represented? One way of doing it is to try and develop different types of non-monotonic logics, logics for ‘jumping to conclusions’, ‘scripts’, etc. The other way would involve taking (human) learning strategies more seriously and study them. As a learning strategy, ignorance appears to presuppose human abilities and capacities explicitly. Not only that. Human ignorance appears also to rely upon other learning strategies for it to work well: strategies that help you, for example, to ‘zero-in’ on something at some specific level of abstraction. Ignorance might help you ‘zoom-in’ on something, while holding ‘the rest’ stable. Some other learning strategy, I think, helps you in directing the focus, instructing you where and how to ‘zoom in’. In either case, it should be interesting to wait and see whether and how they will end up ‘representing’ ignorance explicitly.
We have discovered that ignorance (as absence of knowledge) is a precondition for the acquisition of knowledge; I have accounted for it by showing how ignorance (as a learning strategy) creates the preconditions for learning. These preconditions, at times, also hinder the acquisition of knowledge: that is because inducing identification of the two structures also entails the production of ad hoc hypothesis now and then. In other words, it appears to me, my account captures the insights from both traditions without trivialising either. Not only that. By reformulating the problem in terms of a learning strategy, research can now be carried out in a way it could not be before.
10.1 As I have noted before, ignorance is translatable as ‘not-knowledge’ (at times, also as ‘not-learned’, as avidya) in Sanskrit. Let me take the Buddhist tradition as an illustration to make one more point in this context. The basic insight of meditative practices is that structures like ‘self’, ‘agency’, etc. are not to be found in experience itself. Rather, they are the structures provided by descriptions. That is to say, the insight involves in dissociating experience from its explanation. What ignorance does is to induce identification between these two structures, something that meditation pulls apart. However, the question here is: why are structures like ‘self’, ‘agency’, etc. less easily susceptible to ‘criticisms’ than others? I will take up this question soon.
10.2 Even though I have used Lakatos and Neurath in the course of this article, it would be foolish to identify the strategy of ignorance with the strategies used in scientific problem-solving. Some general strategies inherited from evolution, some cognitive strategies human beings use in their cognitive activities will also be used in scientific theorising. But it does not follow from these uses that scientific theorising is the set of such strategies. I still think that my characterisation of science and scientific knowledge (as I formulate it in ‘The Heathen …’) is at an appropriate level: it is a culturally specific knowledge. It will also have evolved cognitive strategies specific to itself which are learnable by any human being, but have emerged in a specific configuration of learning. In that sense, what Lakatos has identified is not specific to scientific research programmes but applicable to all human learning. It is important to keep this distinction firmly in mind.
11.1 What then is experience? In the most general terms, it refers to a relational attitude in the world towards all objects, events, actions and such like. (I am trying to conceptualise the Sanskrit word ‘Anubhava’ here.) Ignorance enables such a relational attitude by holding either the experiencer or the experienced steady and constant. However, ignorance also does more: it masks the relational attitude by inducing identification between one of the relata (the experiencer and the experienced are the relata in our case) and the relational attitude. That is to say, ignorance masks experience. This seems to be a paradoxical claim: a learning strategy immunises the precondition of experience by masking experience. In other words, one could have experience if and only if one does not have experience! Or, one could have a relational attitude if and only if one does not have a relational attitude. I think this has only the appearance of a paradox. One speaks about experience as though it is a property of either the experiencer or the experienced. I ‘experience’ something in some way either because it has to do with me or with the world. What disappears from the picture (in this way of talking) is the relational attitude even though such an attitude is present.
Is this way of speaking intrinsically wrong? This is an empirical question that cannot be answered before investigation. Why? In many cases, we can eliminate a relationship between two objects by speaking of them as ‘dispositional properties’ of the objects in question. The facts that sugar dissolves in water and a magnet attracts iron are dispositional properties of sugar and magnet respectively: one is solvable, whereas the other creates a magnetic field. In other words, we can rightly conceptualise relations as dispositional properties of organisms or as causal forces of Nature. There is nothing intrinsically wrong with this. Therefore, there cannot be anything intrinsically wrong with what ignorance does. This shows that ignorance could indeed be a result of our evolutionary inheritance.
11.2 Because transforming relations into dispositional properties is a way of conceptualising a relationship, the learning strategy that ignorance is ends up masking the relational attitude that experience is. This masking is possible because of the conceptual scheme we use. This conceptual scheme identifies the nature of the relational attitude with either the experiencer or the experienced. The identification of experience with either of the two relata makes not only conceptual but also evolutionary sense. In doing so, ignorance as a learning strategy immunises the precondition of experience.
Now we can try and make sense of some of the earlier imageries and formulations. ‘Experience is an absolutely structure-hugging fabric’ was one such imagery. The relational attitude can hug either the experiencer or the experienced absolutely. ‘An induced identification between the structure of experience and the structure of the cognitive scheme occurs’ was a second claim. This identification is possible because the relation between two objects could be conceptualised as the dispositional properties of the objects in question. However, this conceptualisation does not make the relation disappear; it is only a way of understanding that relationship. The presence of the experiencer and the experienced is the absolute precondition for experience. Ignorance immunises them both by masking the nature of experience. The constancy and stability of the relational attitude is enabled by identifying the attitude with either of the two relata. In short, ignorance works. It is a precondition (in this sense as well) for knowledge.
How does our notion of enlightenment fit into this picture? Pretty neatly, would be my first guess. (What follows is more speculative than what has gone before.) The enlightened does not conceptualise the experience as the dispositional property of the agent or as the dispositional property of the object. His experience is not masked. He has broken the shackles of ignorance by splitting the nature of his experience from its conceptualisation. To do this, he has learnt to dissociate the emotions that induce this identification. He is the experiencer because he has a relation to the experienced. After all, experience relates him to the world. What is the nature of this relationship, if he does not conceptualise it? He has “Anu- bhaava” and not merely “Anu-bhava”, ie., he has the appropriate (Anu) ‘feeling’ (bhaava) and not merely an appropriate (anu) existence (bhava). He describes this attitude and reflects about them: these are the so-called teachings from the enlightened. These teachings teach you to have the same ‘bhaava’ and that is how they can guide us. He is not ‘one’ with the experienced (this idea does not make sense outside the religious mystics in the west); nor is he the ‘other’ of the experienced. The experiencer and the experienced are related to each other the way the two poles of a magnet are connected to each other. They are indivisible: they are one as a magnet; two as poles; united through the relationship.
I do not know whether I have made much sense. But this should indicate the direction I want to travel when understanding the Indian traditions.