In David Hume’s An Enquiry Concerning Human Understanding, Hume discusses analytic propositions and synthetic propositions, which he calls “relations of ideas” and “matters of fact”. According to Hume, the veracity of an analytic proposition can be verified simply by consulting our ideas, or the constituent words of a sentence, but a synthetic proposition is determined by facts of our experience. If the veracity of a synthetic proposition is causally related to our experience, how then, Hume asks, can we be justified in supposing the veracity of synthetic propositions of which we have no experience? The only possible way to do this, says Hume, is to assume the uniformity of nature, but this assumption is not justified; when making predictions, our conclusions about the future are not a logical consequence of the evidence. Nelson Goodman, in Fact, Fiction, and Forecast argues that this normative problem has been dissolved, and he offers a new riddle. I hope to show not only that Goodman’s supposed dissolution of Hume’s problem has failed, but also that Hume mistakenly accepted the inevitability of induction. More specifically, I will argue that our reasoning is not inductive, but that it is conjectural and hypothetical; our theories are grounded in universal cognitive principles crafted by natural selection and falsified by our observations. I will call this method of reasoning “speculative conjecturalism”.
The Grue Paradox
Nelson Goodman, unlike Hume, rejects the proposition that all regularities are confirmed by their instances. He instead argues for two types of regularities: law-like regularities that are confirmed by their instances, and non-law-like regularities – contingent or accidental (Goodman, Fact, Fiction, and Forecast, 74). He does so by reframing the original enumerative induction argument, and inserting a new predicate. The original argument can be summarized as follows:
All emeralds observed thusfar are green, therefore, all emeralds are green.
To illustrate the inductive invalidity of the above argument, Goodman offers us the Grue Paradox, which includes his new predicate:
All emeralds observed thusfar are grue, therefore, all emeralds are grue.
In Goodman’s example, an object is grue iff it is either green, and observed before time t, or blue, and not observed before time t. The particular time chosen is irrelevant; any time set in the future will do.
According to Goodman, his new riddle illustrates that the statement “all emeralds are grue” is not confirmed by its instances. More importantly, it shows that Hume’s assertion that induction is unable to support any synthetic propositions is actually incorrect, and that the opposite is true: induction can support a potentially infinite amount of synthetic propositions. For any one piece of evidence, there are an infinite number of competing theories, all inductively supported – any induction confirms any induction. Thus, Goodman says, the problem of induction remains unsolved. Of course, there are objections to the grue paradox, some of which were anticipated by Goodman. For example, in Fact, Fiction, and Forecast, he admits that some might wish to define law-like regularities according to their generality content, or to reject certain kinds of predicates. Goodman and others, for example, Hillary Putnam, have been mostly successful in rebutting attacks of these kinds, and I will not summarize their defenses here. Instead, I will focus on one particular defense of the use of grue-like predicates, namely that a principle of simplicity does not sufficiently dispel them, and I will argue that the question of entrenchment is irrelevant, since reason is not contingent on language, and it is not inductive.
Universal Cognitive Principles and the Discontinuous Mind
As Jerold J. Abrams relates in his essay “Solution to the Problem of Induction: Peirce, Apel, and Goodman on the Grue Paradox”, Hillary Putnam has, in defense of grue-like predicates, argued that any objection to the use of grue-like predicates “…is inadequate because predicates are relational to the language in which they are used” (Abrams, 545) and to this I agree. The argument that the predicate “green” is simpler and therefore more favorable than “grue” is untenable because it begs the question – we already prefer “green” to “grue,” and it is from this vantage point that we dismiss the subjectively evaluated “complexity” of “grue”. In this case, Occam’s razor is of little use, that is, unless there exists some reason to believe that “grue” is objectively more complex than “green”. If there exists some evidence for human cognition to prefer green-like predicates to grue-like predicates, than we might be justified in concluding that grue-like predicates are unnecessarily complex, and discard them in accordance with Occam’s razor. I hope to show that an objective checksum for cognition does exist, namely, that universal cognitive principles exist, and that these principles explain why we prefer “green” to “grue”.
Nathan Stemmer, a psychologist, and author of “Hume’s Solution of the Goodman Paradox”, calls universal cognitive principles “innately based inductive inferences” and tells us these inferences had “…survival value” (Stemmer, 16). True to evolution, Stemmer says, our modes of inference that were unsuccessful or less “fit” were lost, while more fit methods remained. This hypothesis has, at least in part, been confirmed. In “Evolution of Consciousness”, John C. Eccles says, “…consciousness appears to have come into the mindless world of biological evolution with the origin of mammals” (Eccles, 7323). He goes on to say that current evolutionary theories give much evidence for the conclusion that “conscious experience would give evolutionary advantage”. Though Eccles admits that the research concerning Homo sapiens sapiens is as of yet inconclusive, he does say that we now “… have an evolutionary explanation for the phylogenetic origin of mammalian consciousness. It would appear initially in the primitive cerebral cortices of evolving mammals, such as the basal insectivores of today” (7323). Alejandro Gonzalez-Voyer, Svante Winberg and Niclas Kolm, in “Distinct Evolutionary Patterns of Brain and Body Size during Adaptive Radiation”, give us yet more evidence that evolution is directly responsible for the attributes of the brain: “…brain evolution is thought to be influenced by the biotic and abiotic environment of a species. Indeed, brain size has been shown to be related to ecological factors, such as habitat (Safi and Dechmann 2005) and diet (Gonzalez-Voyer et. al. 2009)” (2267).
The fact that there is a causal link between our environment and our cognition seems undeniable; natural selection has crafted the biological functions of our brains in the same way that it crafted the rest of our body. This is further demonstrated by the fact that this phenomenon has occurred among other animals. For example, according to the Stanford Encyclopedia of Philosophy, other animals have developed IBE as a result of evolution (Allen, Colin, “Animal Consciousness”, The Stanford Encyclopedia of Philosophy). Since natural selection has causally acted upon our brains and has shaped our consciousness, it seems natural to assume that universal cognitive principles would develop, since natural selection would favor more fit cognitive functions. Indeed, this is just what we find; there is a commonality of reasoning among Homo sapiens sapiens. One such commonality is a phenomenon called the discontinuous mind, and it is this phenomenon that allows us to reject grue-like predicates.
Richard Dawkins, in an article for “New Statesman,” describes the discontinuous mind as a mind that creates “…a gratuitously manufactured discontinuity in a continuous reality” (Dawkins, 2011). This discontinuous mind, or the craving for absolutes, creates in humans a desire to categorize and subcategorize everything they encounter – black or white, yes or no, hot or cold. What humans often fail to realize is that all of these categories are subjective. Indeed, there is no existential hot or cold, nor is there an existential black or white – our subjective labels for properties of reality are entirely arbitrary. Nevertheless, categorizations have been indispensible tools for our species. Useful cognitive principles would likely never have developed, had we lacked our hard-wired tendency to categorize. How could our ancestors decide which way to turn, what to eat or what not to eat, what was hot or what was cold, what was light, or what was dark – without any ability to parse the continuous? The discontinuous mind allows us to break the continuous world in smaller, more manageable parts, and it exists necessarily – crafted, like all of our other universal cognitive principles, by natural selection, and it is the discontinuous mind that repels from Goodman’s use of “grue”.
Rejecting a grue-like predicate because it is more complex than our currently accepted predicates is circular, but rejecting a predicate because it conflicts with our universal cognitive principles is not; it is by no subjective opinion or analysis that we reject “grue”. Rather, grue-like predicates are discarded automatically according to our psychological programming and inherited universal cognitive principles. Natural selection is to blame for our conclusion that grue-like predicates are unsavory. An objective check sum exists within our psyche that has been designed to abandon predicates which are overly complex, or continuous. Occam’s razor may not apply in the sense of subjective simplicity, but the discontinuous mind objectively programs us to the same conclusion. “Grue,” with its combination of both qualitative and locational predicates naturally appears to the brain to be more continuous, and more complex, whereas green sits well with our psyche, since it is merely qualitative. Thanks to natural selection and evolution, we have an objective means for determining the undue complexity of “grue”.
Beyond the issue of unsavory predicates, Goodman presents us with an entirely arbitrary and subjective argument. Even if grue-like predicates are accepted, we are left with a method of reasoning that avoids objectivity entirely, since an inference to Goodman works whether or not it is true, so long as it is entrenched. His argument suggests that induction is contingent upon language, rather than language being contingent upon induction. This makes little sense. Would not our reasoning come first, and our language issue forth from our reasoning faculties? How could predicates be meaningful for one of our ancestors, who lacked language, but nevertheless reasoned? Moreover, induction ought to be a truth-seeking enterprise, but Goodman instead offers a completely subjective form of reasoning that is unconcerned with the truth. In the end, his philosophy is far too arbitrary to be useful. Lastly, his philosophy rests on a faulty foundation – the assumption that induction exists. If induction is false, as I believe, then the nail will be driven through Goodman’s proverbial coffin.
Proponents of the induction myth would have us believe that we reason from particular instances to generalizations, but this assertion seems unnecessarily linear and unfounded. In Hume’s example of the billiard balls, we are to believe that a person with no prior experiences, who saw one ball moving toward another would have no guess as to what would happen when the balls collided. Perhaps the balls would explode. Perhaps one ball would pass through another. Without any prior experience, says Hume, we would have no way of predicting the outcome, since we would not be able to create a generalization from particular instances. Goodman would have us believe that, should we succeed in any inductive inference, we would be unjustified, since we would have absolutely no predicates at our disposal. Is this the case? Would we be unable to predict anything concerning the billiard balls, if we had no prior experience and no predicate entrenchment?
As I have argued above, human beings have inherited, from natural selection, certain universal cognitive principles. As the psychologist Roger Shepard has pointed out in a recent lecture, our mind is “optimized for a three-dimensional Euclidean world” (Shepard, “Principles of Cognition as Adaptions to the World). An innate method of reasoning resides within us – passed from generation to generation, over thousands of years. Principles such as the discontinuous mind, the anthropomorphic tendency, the actualizing tendency, and others make up the foundation of our psyche. These principles causally act upon our conscious mind, before we are even aware of what we are thinking. In essence, the mind is deterministic and holistic. As Sam Harris pointed out to his readers, in his book Free Will, “You are not in control of your mind – because you, as a conscious agent, are only part of your mind, living at the mercy of other parts. You can do what you decide to do – but you cannot decide what you will decide to do” (Harris, 38). If someone with no prior experiences suddenly became sentient, they just might guess what would happen to a pair of billiard balls.
Just as a salmon needs no prior observation or experience to know that it must swim upstream, and just as dog needs no prior experience to know that it must suckle at its mothers bosom, so too does the human mind gift us with innate cognitive principles. The argument for what qualifies as instinct and what does not is, though interesting, irrelevant. In every case, instinct or not, we are presented with the obvious fact that processes of the mind occur according to the makeup of the mind itself, not at the behest of our conscious will. How then, if not inductively, if not from instances to generalization, do we reason? The answer is in speculative conjecturalism.
In every case that we predict something about the future, we do not do so based upon instances of particular observations, or necessarily on any observation at all. Rather, we conceptualize a particular scenario speculatively, and we conjecture as to what might happen. If the mind has no observations from which to drawn on, then the universal cognitive principles of the mind will provide a foundation from which to form hypotheses. In the case of the billiard balls, our minds, over the course of evolution, have been formed to expect basic properties of physics, such as one thing colliding into another – in much the same way that birds have been cognitively sculpted to build a nest. This is no guarantee, of course, that we would guess correctly. Perhaps our intuitive reaction might lead us to believe that balls would explode. The process of speculative conjecture is not at all relevant to the conclusion, to predicates, or anything else, and in the case of the billiard balls, it is impossible to determine immediately which of our potential theories are explanatorily superior. Rather, the theory must be assessed after the fact, as Karl Popper suggested, by falsification. If we supposed that the billiard balls would explode, and instead they merely collided, our first theory would be discarded. This is not an inductive process, because the explanatory power is in the theory itself. We do not suppose that one theory that has worked in the past will continue to work in the future, and we do not derive our knowledge from repeated observations. Our observations merely serve to falsify our conjectural knowledge. In a sense then, Popper was right: we cannot arrive at any conclusion without falsification. This makes perfect sense when we consider our own reasoning; any semblance of inquiry must begin with doubt, and to doubt is to conjecture – to suppose that something is false, regardless of any observation.
From a foundation of universal cognitive principles, we speculate and conjecture. We do this by creating a coherent network of beliefs. Our first observation is checked for coherency against our innate principles, and our first theory forms automatically. A second observation is entered into the equation, and we check this observation against our first, and against our checksum of principles. More observations are added, and the coherentist account of knowledge begins to form. Along the way, with each new hypothesis, we guide ourselves by Popperian falsification. The process remains hidden to us because we are part of the process – innate evolutionary patterns of conjectural reasoning are acting on our mind. This hidden process fools us into believing that we are the conscious agents of our thoughts, and that we have developed generalizations based on particular observations, but in reality we form generalizations according to coherence with our network of beliefs, and our innate principles. We speculatively conjecture, based upon our programming – and this type of reasoning is justified, since evolution has proven it successful and necessary. Thus all linear accounts of knowledge can be discarded, the coherentist web can be embraced – its origin explained. Arbitrary theories of knowledge that rely upon inhuman predicates can be discarded.
I am sure there are many objections to this theory, as I am afraid I have not had enough time to explicate my thoughts, and I may have been deficient in my delivery. I am sure, for example, that some may take issue with species-memory habits, instinct, and universal cognitive principles. This is to be expected, though I hope I have referenced well enough a few academic sources that support the idea that we are not the conscious agents of our minds. Beyond this, I cannot respond more to this objection, other than to urge my readers to search for any information available concerning cognition and how it relates to evolutionary biology.
Another objection I foresee is concerning my use of the term “speculative conjecturalism”. How, some might ask, does my philosophy differ from Karl Popper’s falsification theory? The answer lies in the fact that Popper discards induction, arguing that it has no place in science, but nevertheless does not explain fully the sources from which our theories arise – at least, not to my satisfaction. I would like to make the point that in some cases at least, or theories might arise from the void, much like our thoughts arise from the void. We can only explain our thoughts in retrospect, much like we can only establish the usefulness of a theory, in Popper’s opinion, in retrospect. In other words, we do not always formulate hypotheses, because that would require conscious agency. Rather, hypotheses arise from the void, or more precisely, as a web of ideas that reflects our innate and universal cognitive principles. This web gets us into the territory of coherentism, and issue of justification is solved, because our beliefs systems are justified, not by any one belief, as in the case of foundationalism, but by coherence with the system.
Yet another objection might concern my denial of grue-like predicates on the basis that they are objectively unnecessarily complex. If my first premise concerning universal cognitive principles cannot be accepted, my rejection of “grue” is even more tenuous. More importantly, I have rejected grue-like predicates under the assumption that compound predicates that are both qualitative and locational (or qualitative, quantitative and locational, or any unnecessarily complex combination) conflict with the discontinuous mind. One might make the case that there is some universal cognitive principle that prefers grue-like predicates, however, I see this objection as baseless, since it is fairly unanimously agreed among the philosophers who have addressed Goodman that their first inclination was to reject grue-like predicates. The question of why people find grue-like predicates unsavory is, I hope, answered in my discussion of the discontinuous mind.
Some might argue that I have performed an inference to the best explanation, concerning my acceptance of evolutionary theory. I would like to point out that this is not the case, but, like Popper, I find the theory of evolution highly corroborated, and I find that the theory’s explanatory power resides in and of itself. One piece of evidence – a rabbit skeleton in a Cambrian stratum, for example – could falsify the evolutionary theory in an instant. I have not implemented IBE, nor have I made an inductive inference and concluded that the theory will work in the future as it has in the past.
I have argued that Goodman’s philosophy is arbitrary and subjective, that it defeats the purpose of reasoning and abandons truth, that its premise that reasoning depends upon language is false, and that the “grue” paradox contains unnecessarily complex and unnatural predicates. I have also argued that induction, in the sense of forming generalizations from particular instances or from “entrenchment”, is false, and that human beings reason according to speculative conjecturalism – a process that is deterministic and holistic, and founded in universal cognitive principles that have been crafted by natural selection. I have argued that some predictions about the future simply arise in the mind, as a web of beliefs that represent the matrix of our cognitive principles, and that all observations are set against the checksum of these cognitive principles. I have also argued for coherentism, and the idea that the justification for our beliefs is based on the coherence of a system, rather than on a prior belief, but that coherent webs do have a source – universal cognitive principles. If what I have said is at all logical, then we are not the conscious agents of our thoughts, though we are deluded into thinking we are – but we are justified in our generalizations.
Hume, David, An Enquiry concerning Human Understanding.
Goodman, Nelson. Fact, Fiction, and Forecast.
Abrams, Jerold J., “Solution to the Problem of Induction: Peirce, Apel, and Goodman on the Grue Paradox”
Stemmer, Nathan, “Hume’s Solution to the Goodman Paradox”
Eccles, John C., “Evolution of Consciousness”
Alejandro Gonzalez-Voyer, Svante Winberg and Niclas Kolm, “Distinct Evolutionary Patterns of Brain and Body Size during Adaptive Radiation”
Allen, Colin, “Animal Consciousness”, The Stanford Encyclopedia of Philosophy
Dawkins, Richard, New Statesman, “The Tyranny of the Discontinuous Mind”
Shepard, Roger, “Principles of Cognition as Adaptions to the World”
Harris, Sam, Free Will