Greg Detre
Sunday, November 18, 2001
Thomas Nagel wants to defend reason as providing authority and a means of persuading others of the relative merits and demerits of our beliefs, and so to answer the question �How is it that contingent, biological creatures such as ourselves can have access to such universally valid methods of objective thought?� He wants to deny the subjectivist who thinks that �the first person, singular or plural, is hiding at the bottom of everything we say or think�, or indeed the relativist who wants to deny that there is any objective reason to prefer some beliefs over others.
Reason then, for Nagel, encompasses any means at our disposal of accessing, recognising and producing knowledge that is objective. Our aim as thinkers and rational agents is to arrive at principles that are �universal and exceptionless� � to be able to come up with reasons that apply in all relevantly similar situations, and to have reasons of similar generality that tell us when situations are relevantly similar.
fit in quote from tVfN about forming
representations about ourselves within the world, which we then see as itself just
another
differentiate between objectivity
and universality
This is the strongest objectivism about reason that I am going to consider. It is the most interesting form because while its acceptance would valuably shore up the philosophical enterprise epistemologically against a number of deep skeptical attacks, it still allows for our finitude and fallibility as reasoners, as it must. There are stronger forms of rational objectivism, leading all the way up a God-like form of rationality that admits infallible access to and production of all objective knowledge. There are weaker types that form a continuum towards a relativist, Darwinist sort of pragmatism, where our most rational thought processes have evolved (despite perhaps not even being particularly adaptive) to produce beliefs relativised to our ecological niche x0,000 years ago.
Before going any further, I�d like to demolish any simplistic assertion that we are unequivocally, wholeheartedly rational by putting forward some of the strongest evidence and examples of human irrationality. Any theory of rationality must eventually explain these.
There are various anecdotes and experiments from psychology, especially developmental psychology, that point towards our minds as developing over time to fit the niche we occupy in the universe. For example, children insist that a grain of rice weighs nothing, while a bag of rice is heavy. Young children fail the false belief test, assuming that people know the outcome of events they did not witness or hear about. Of course, we know that those children become more rational over time � but it is worth remembering that at least at an early stage in our lives, humans are manifestly irrational in almost every domain, reiterating a picture of rationality as graded, at least to some degree, which I will discuss later with regard to Cherniak and �minimal rationality�.
As Pinker puts it, these children �grow into adults who think that a ball flying out of a spiral tube will continue in a spiral path and that a string of heads makes a coin more likely to land tails�. The former mistake is perhaps not so much one of rationality, but of an imperfect model of the physical world. What is more interesting about this mistake is that when the same people (including college students) are shown an animation of a ball flying out of a spiral tube in a spiral trajectory, they burst out laughing. This illustrates a second important point � even when our intuitions or initial judgements are wrong, with a little help our errors sometimes become easily apparent.
The issues surrounding probability are more curious still. Tversky and Kahneman provide plenty of examples of people�s intuitions about probability leading them systematically astray, e.g.:
It seems that the brain is adopting rules of thumb in place of theorems, e.g. the more memorable an event, the more likely it is to happen (plane crashes). Of course, such behaviour can often be understood to some degree in other ways. Gambling is a thrill, and plane crashes are off-puttingly horrific.
People�s
performance at falsifying hypotheses is also mixed. Wason told subjects that a
set of cards had letters on one side and numbers on the other, and asked them
to test the rule �If a card has a D
on one side, it has a 3 on the
other�, a simple P-implies-Q statement. The subjects were shown four cards and
were asked which ones they would have to turn over to see if the rule was
true.�
������ D��� F���� 3���� 7
�Most people choose either the D card or the D card and the 3 card. The correct answer is D and 7. �P implies Q� is false only if P is true and Q is false.� Only about 5-10% get it right. It�s not because people assume it�s an �iff� statement, otherwise they�d turn over all the cards. People seem to be �confirming their prejudices rather than seeking evidence that could falsify them�. Cosmides has found that rephrasing the experiment with real-world events helps, e.g. phrasing a logically-identical problem in terms of a bouncer checking for under-age drinking � but only when the rule is a contract, an exchange of benefits. She cites this sensitivity as support for Trivers� prediction that humans, as �the most conspicuous altruists in the animal kingdom, should have evolved a hypertrophied cheater-detector algorithm�.
The discussions about probability and falsifying hypotheses serve to make two more points. They show that our thought processes seem to be biased towards having or forming beliefs and habits of thought that are useful, efficient, approximate the truth and are perhaps specialised for the world in which we evolved. Lastly, we sometimes seem able to employ higher-order processes (involving language, abstraction/formalisation and step-by-step breakdowns) to validate and quantify these semi-intuitive conclusions and form the sort of objective frameworks that exemplify Nagel�s conception of rationality. Indeed, this is how we are able to step back and devise the mathematics of probability or mechanics to see our error.
However, the fact remains that there are discrepancies between the way we naturally think and the careful, academic, rational (e.g. mathematical) way. I will argue that these discrepancies result from having evolved parochial inference mechanisms which exploit long-term regularities in our world, since real science and a generalised rational approach are expensive and slow, and because evolution shapes specifically for fitness rather than truth. The questions that remain to answer are whether our inferential mechanisms are sufficiently powerful in their generalising to qualify as rational, and if so, to what extent this is explicable in terms of a naturalistic picture.
I now want to move away briefly from empirical issues for rational objectivism to a few more general concerns about the limitations of our rational capacity that a theory of rationality would have to incorporate.
phenomenal character of rationality
Firstly, we continually find that some concepts, arguments or ideas are more complicated, remote or difficult to grasp, follow or apply than others. I am not referring to the trivial fact that we have imperfect senses and finite, fading long-term memory. I am not really referring either to differences in deeper cognitive capacities, such as limitations on the complexity of syntax we can parse, levels of articulacy, varying success with arithmetic or abilility to imagine a scene. Rather, I am talking about fundamental restrictions and variations between how easy thoughts are to conceptualise, convey or fully absorb, or certain processes to follow and criticise. What we can conceive is bounded. This is so obvious to us that we scarcely consider its significance. Similarly, it is trivial to point out that different people�s abilities in these areas vary, and even a single person�s faculties wax and wane with diet, hours of sunshine in the day or stress.
It is an ineliminable aspect of our description of an omniscient being that it would find all such conceptual hurdles to be inconsequential. We cannot usefully speculate over whether such an infinite mind would consider all problems and concepts equally trivial, or whether it would recognise some to be harder than others but still confidently within its ability. Either way, the fact that we falter, and sometimes fail entirely, to understand new things is clear evidence that our rationality is bounded.
Furthermore, our bounded rationality is evidenced every time two intelligent, humble and well-meaning, sane and highly rational experts seeking a resolution on their topic meet to debate, and disagree. Sometimes, one convinces the other. Perhaps they will realise that they agree after all, and their disagreement was based on a minor definitional point. Often, especially in science, there simply isn�t enough data available to incontrovertibly confirm their respective hunches. Occasionally, they agree to disagree because the resolution comes down to an article of pre-rational faith, such as religious beliefs.
The charitable assumption implicit in these explanations of why the experts� (who will label philosophers, since they make excellent case studies here) views are so often irreconcilable is that they simply have different �epistemic profiles�. This is a handily vague term that stretches over our entire framework of beliefs and desires, and masks a multitude of sins. It is intended to describe the ease or degree to which one accepts things as justified, or the nature of justification one demands. Different people might be more or less predisposed to employ formal logic, mathematics, a reasoned argument or speech, intuitions etc. as a means of justification. When our two philosophers continue to disagree then, it is because the sorts of justification they look for are different. One might regard the syllogism as mightier than the scientist, or have a penchant for the slightly mystical (e.g. panpsychism, belief in God). It might be that their inability to agree on the issue at hand rests on a long-forgotten willingness in their undergraduate days to adopt certain invisible background premises which have come to seem self-evident. One might point to the very small proportion of professional philosophers whose views have changed substantially in print over the course of their careers as evidence of this.
Is it evidence of a fundamental human non-rationality that intelligent, truth-seeking philosophers have differing epistemic profiles?
EP = what they require in order to feel certain about something, what they prioritise, favoured methods, indoctrination, pre-rational faith (e.g. religion)
one can still be rational within one�s epistemic profile � and indeed, especially if we accept Nozick�s optimism about reason�s capacity for self-analysis and �repair, then we can examine and adjust our epistemic profile
what if it�s not just their epistemic profiles, but that they�re actually just making mistakes/fallacies in their reasoning
Ascribing their discrepancy of opinion to epistemic profiles is more charitable, and leaves more room for rationality than does the alternative � we constantly, unwillingly and unwittingly err in our reasoning. We pick these errors up regularly, or others do for us, but it seems quite possible that however powerful in some ways our reasoning capacity is, these are symptoms of what is at heart a jury-rigged composite of souped-up survival-enhancing behaviour modules. These fallacies take the form of mis-applications of definitions, erroneous steps in an argument etc.
In order to understand how these arise, account for them and discern whether they threaten our entire claim to rationality, a theory of rationality may need to delve deep into the sub-cognitive, even low-level neural implementation of our thought processes. Later, I will try to consider whether what we do know about the brain is compatible with Nagel�s conception of rationality.
Altogether, these obstacles to clear thought lead me to an unshakeable wariness about our level of rationality in general.
Of course, all of this discussion is particularly subject to Nagel�s argument against any attempt to subjectivise reason that such attempts are inherently self-defeating. I will address this later.
It seems clear to me that Nagel�s priority and motivation for holding his position is his faith in a reason that can help us apprehend the objective (and keenness to justify that faith);
mention epistemic profiles
mention Nagel�s (good) reasons for wanting
to preserve faith in reason?
on the other hand, my own corresponding prejudicial, prerational or intuitive faith (that I would like to feel justified in having) is in a scientific understanding of the workings and origin of humanity. At the same time, we would both like to be able to incorporate the opposing position, which is problematic for our starting tenet but which we consider to be otherwise desirable and fairly plausible. In Nagel�s case, he would like to try to reconcile naturalism (e.g. an expanded Darwinism) with his objectivism about reason. In my case, I would like to try to reconcile a similar sort of rational objectivism with my naturalist tendencies. Explain briefly why naturalism and rational objectivism are difficult to bring together. However, if Darwinism can�t be expanded to show how humans have evolved to be rational, then Nagel is content to say that the burden of proof lies with the Darwinists, and until then we should consider it a handy but provisional and incomplete representation of the universe that we may never fully perfect. And I want to say that if the rational objectivist cannot show how such an objectivity-tracking cognitive capacity could have evolved, then we simply cannot be (and we are not) rational in the way that Nagel argues.
Of course, I don�t think that Nagel would appreciate having his rational objectivism branded a prerational faith, and I don�t think I�m being fair to myself either. This way of putting things is mainly intended to emphasise that epistemological discussions are so crucial, and so entanglingly difficult, that it seems to me that much of the fervour with which objectivist and subjectivist positions are defended is testament to philosophers� own predilections and epistemic constitutions more than anything else.
I have deliberately couched the naturalistic account given above in terms of a late twentieth century understanding of homo sapiens and the universe. Such a world-view is defined by its neo-Darwinist account of our origins, and its aim of a complete picture of the workings of our mind in terms of cognitive psychology, ultimately fully translatable into (and reducible to) brain states. Of course, the picture available at the end of the twentieth century is not adequate to explain any high-level cognitive processes in psychological, let alone neurophysiological, terms. However, such problems as understanding in physical terms how we learn or use language or do simple arithmetic are �easy� cognitive problems (and form an inherent part of any naturalistic account) relative to what might be termed �hard problem of rationality� of understanding how our thought processes map onto an objective framework. The problem is �hard� because even investigating it requires one to have already solved it. It is this hard problem which does not sit well with naturalism.
it is simply the problem of objective knowledge - �Objectivity itself leads to the recognition that its own capacities are probably limited, since in us it is a human faculty and we are conspicuously finite beings. The radical form of this recognition is philosophical skepticism, in which the objective standpoint undermines itself by the same procedures it uses to call into question the preflective standpoint of ordinary life in perception, desire and action. Skepticism is radical doubt about the possibility of reaching any kind of knowledge, freedom or ethical truth, given our containment in the world and the impossibility of creating ourselves from scratch.� �In general, I believe that skepticism is revealing and not refutable, but that it does not vitiate the pursuit of objectivity.� (tVfN)
I am not going to define the naturalistic account in terms of the contents of twentieth century journals, because it is clear that such a characterisation is more handicapping, restrictive, parochial and defeasible than the naturalistic account needs to be. Even so, late twentieth century science is the most attractive naturalistic account available at the moment, and it�s convincing enough for it to have become epistemically foundational for me.
Rather, borrowing from Rorty, I would like to define the naturalistic account in terms of what would be acceptable to an ideal scientific community in the future. An �ideal scientific community� comprises educated, sophisticated, intelligent, open-minded but critical scientists, like today�s scientists at their best, working with the benefit of the progress made during the intervening years accrued purely through incremental science.
The problem with this way of defining the naturalistic account is that it rests too heavily on just how distant a future we are talking about. Words like �accrued� and �incremental� conjure an image of scientists gathering nut-sized nuggets of data assiduously into ever-growing piles. The problem I am most concerned about with regard to this image is the idea that each increment of progress is bite-sized, i.e. supports and builds upon previous theory linearly. In reality, the march of science proceeds in fits and starts, although the intervening periods may be stable and consolidatory. I�m worried that just one or two paradigm shifts in succession (on the scale of the theories of evolution or relativity, for instance) are enough to distance earlier and later scientific viewpoints almost unrecognisably. A sufficiently futuristic naturalistic account may be so different, employ a barely commensurable vocabulary, and shaped to fit a quite different cultural context, that it somehow renders the current debate about reason�s objectivity obsolete.
In order to form then a concrete naturalistic opponent for the rational objectivist, I am going to plump for a naturalistic account defined in terms of what would be acceptable to an ideal scientific community in the not-so-distant future. This way, if we do decide in the end that a naturalistic account defined so squarely within a twentieth century framework is not compatible with reason�s objectivity, and at the same time have not dislodged the objectivist�s arguments, then we will not simply be left in an unhelpful, rigid stalemate between rational objectivism on the one hand, and naturalism on the other. Instead, we can probably say that we over-hamstrung the naturalist by wedding him to a framework which is highly likely to expand and flex considerably. The next step would be to try and relax our characterisation of the naturalistic account (at the expense of vagueness) to see whether it might be more accommodating to rational objectivism, a task that inevitably benefits from actual scientific advances to measure itself against. We might hope that we can even pinpoint which particular areas of a near-future naturalistic account are most at odds with rational objectivism, and for what reasons.
Of course, I would be happiest if it were possible to find a means of accommodating the rational objectivist within some turboboosted, but recognisable, version of the twentieth-century scientific framework.
This leaves the contentious question of how exactly to characterise the twentieth-century framework. I have tried to choose a minimal set of principles to embody the late twentieth-century scientific picture, all of which have a direct bearing on our understanding of the rational process, and which I consider to be explanatorily and predictively generative and coherent, and elegant:
philosophy of mind
physicalism
functionalism
extent to which consciousness plays a role in reasoning
neurophysiology
connectionism
super-Darwinism
psychology
society
development
evidence (include here???)
Such a list of principles is an unavoidably subjective task. Those who take issue with my choices can see this as another point from which to rally the naturalist�s cause if necessary.
I should also explain why I have so far ignored the relevance of consciousness to our rational processes. It might seem intuitively obvious to some that consciousness is intimately related to rationality, while it might seem intuitively obvious to others that the opposite is true. As I will outline below, I have deliberately chosen a materialist stance in philosophy of mind in my naturalistic principles, which allows me to avoid treating consciousness as a separate phenomenon, but rather assume it to be incorporated into related discussions of neurophysiology etc. This is partly because I�m keen to drive as much mystery out of the brain as possible before beginning to discuss rationality (especially given Nagel�s stance towards naturalism, or �scientism� as he derisively terms its extreme forms), and partly because materialism is an entrenched dogma of (late) twentieth century (cognitive) science.
I will come back to discuss each of these principles in more detail, starting with evolution(???).
consider this near-future naturalistic account as �scientism(???)�, a form of idealism
incorporate cherniak, �naturalised
epistemology� and the kornblith introduction
important question � to what extent, even if our rationality isn�t fully objective, can it become more so (cf Nozick)???
�I
have to be able to believe that the evolutionary explanation is consistent with
the proposition that I follow the rules of logic because they are correct � not
merely because I am biologically
programmed to do so�. is Nagel importantly missing the point about �I� and
�programmed� here???