Philosophy 437 Lecture Notes

Daniel Dennett

(Based upon Darwin's Dangerous Idea)

I. Dennett's Philosophical Program

A. Substituting Cranes for Skyhooks

1. A skyhook is an imaginary device that is attached to the sky for the purpose of lifting or suspending objects—an imaginary device because it is impossible; a crane, on the other hand, is an actual device that, with adequate design, is useful for lifting or suspending objects.

2. According to Dennett, many philosophers and others mistakenly depend upon, and even desire, skyhooks as a fundamental means of explanation.

a. The skyhooks might include God, the Mind, the soul, a special state of consciousness, or a simply autonomous self.

b. As a need, philosophers may view skyhooks as fundamentally necessary for explanation—for example, Locke's insistence on Mind (or God--RY) as necessary for the explanation of design in nature.

c. As a desire, philosophers may seek skyhooks as protection against a reductionistic scientific materialism or against the subjugation of the humanities to an indifferent science.

3. Dennett is a skyhook slayer—intent on showing that skyhooks are neither necessary for explanations nor especially desirable for satisfaction of humane goals.

a. Dennett does not say that skyhooks have never served a useful function; he just views them as outdated and he sees the advancement of knowledge as better served through the search for "cranes."

b. In Darwin's Dangerous Idea, he is concerned with finding cranes that explain the design of the evolutionary process.

c. He also stresses that evolution depends upon lots of little cranes, not just big ones and that little cranes are capable though of producing big ones.

B. Taking Darwin's Dangerous Idea Seriously

1. Evolution by Natural Selection as his dangerous idea (p. 42): "But if the variations useful to any organic being do occur, assuredly individuals thus characterized will have the best chance of being preserved in the struggle for life; and from the strong principle of inheritance they will tend to produce offspring similarly characterized." [from Origin of Species]

a. Darwin thereby described how a "Nonintelligent Artificer" could produce adaptations in nature over vast amounts of time and proved that many of the intermediate stages for such a process had actually existed. (p. 47)

2. Evolution as an Algorithmic Process as his dangerous idea (p. 59): The algorithmic level—a process with "a set of individually mindless steps succeeding each other without the help of any intelligent supervision"—explains the wonderful diversity of nature.

a. Characteristics of an algorithmic process initially: (1) substrate neutrality, (2) underlying mindlessness, (3) guaranteed results

b. Elimination tournaments as examples of algorithmic processes

3. Darwinism functions as a "universal acid" that corrodes and transforms traditional notions, such as fixed essences.

4. "To put it bluntly but fairly, anyone today who doubts that the variety of life on this planet was produced by a process of evolution is simply ignorant . . ." (p. 46)

a. Darwinism has met numerous challenges and has become even stronger as a system of explanation by overcoming challenges—so much so that the burden or proof lies with those who would oppose it. (p. 47)

5. In addition to showing how Darwin's dangerous idea works in evolutionary biology, Dennett applies the idea more broadly to culture, mind, language, meaning, artificial intelligence, and morality.

C. Reductionism Without Greedy Reductionism

1. Good Reductionism in science turns away from miraculous or mysterious initial explanations and recognizes the interdependence of the sciences—for example, that human beings as mammals are dependent upon principles of biology, and that as mammals they are dependent upon principles of chemistry, etc.

a. Only preposterous reductionists (a very rare or nonexistent breed) want to explain human actions simply in terms of equations in physics.

2. Greedy Reductionists oversimply too much, usually by focusing too much on one big crane rather than seeking out the little cranes necessary for adequate explanation—for example, B. F. Skinner's behaviorism.

a. He does allow some value to greedy reductionism as an initial "nice try" that makes progress by avoiding the complexities of a situation. (p. 396)

D. Reason, Science, and the Pursuit of Truth

1. Dennett is quite comfortable with traditional methods of philosophers with respect to reasoning, scientific methods, and the pursuit of truth.

a. He does express some wariness at the dangers of claiming too much (p. 83).

b. He tries to make a persuasively reasoned case for his positions.

2. There is no philosophy-free science (p. 21).

3. As human beings, we hold truth to be "precious." (p. 22)

4. Rejection of a Reasoned Ground for Faith Separate from Reason and Science

a. He rejects any argument for the view that God is simply a subject of faith legitimately separate from scientific method.

b. He does not see any "reasoned ground for taking faith seriously as a way of getting to the truth." (p. 154)

c. He does not think that sincerity and faith should have standing in the face of obviously contrary empirical evidence. For example, he contrasts the faith and sincerity of some witnesses in a murder trial with a heavy weight of evidence.

II. Evolutionary Design

A. Principle of Accumulation of Design - Given some regularity in nature and vast amounts of time, Natural Selection offers ways of producing more complex design.

1. In effect, it is a long process of R & D (Research and Development); there just isn't any intelligent designer at the start.

2. Mind is the effect of evolutionary design, not the cause.

B. Some Design Cranes

1. Sex - sexual reproduction greatly increases the pace of design change (R & D) compared with asexual reproduction.

2. The Baldwin Effect: When a single organism develops some superior characteristic (a "Good Trick"), other organisms in the group with a genotype closer to the Good Trick organism than other members will start to behave in ways closer to the behavior of the Good Trick organism—so that, over time, the Good Trick characteristic becomes more dominant. (pp. 77-80)

C. Possibilities in Design Space

1. Types of Possibility

a. Logical possibility - whatever is non-contradictory

b. Physical possibility - whatever occurs within physical limits, such as the speed of light

c. Biological possibility - for example, whatever is within biological limits for flight

d. Historical possibility - whatever is within limits imposed by historical events that have occurred—for example, six limbed creatures on earth seem to be precluded by the historical nature of past developments

2. The Library of Mendel - the library of all possible genomes

a. Among the myriad of possibilities, some things will be more accessible than others.

b. Logical and physical possibilities are "clamped," that is, fixed—e.g. gravity.

c. Biological possibilities are more open-ended. While there may well be biological laws (similar to the law for gravitation), we need to be careful that we do not mistake locally centered conditions for universal rules.

D. Actualities in Design Space

1. A necessity for life: autonomous metabolisms and some kind of boundary to separate an organism from everything else in the environment

2. A "very good bet" for life: vision of some sort for a organism to exhibit locomotion

3. There are "forced moves" in design space, in the same way that we expect numbers, or arithmetic, to apply universally, even for extraterrestials.

E. Examining Design Space

1. What Darwin showed was that "Intelligence could be broken into bits so tiny and stupid that they didn't count as intelligence at all" (p. 133), bits that were distributed over the vastness of space and time and were participants in an algorithmic process.

2. Task of Inquiry: To determine how all this design work was done—who or what did it, where and when

3. We need to examine the biological engineering of design space.

a. Example A: A general principle of this engineering - "Any functioning structure carries implicit information about the information about the environment in which its function 'works.'" (p. 197)

b. Example B (an instance of historical possibility): ". . . if double-stranded DNA happens to begin to develop, it opens up opportunities that are dependent on it." (p. 199)

III. Evolution and Culture

A. Culture as a Crane

1. Culture is both a crane and a builder of cranes.

2. It operates much faster than genetic evolution.

B. Memes (a concept originated by Richard Dawkins)

1. As genes are the carriers of biological evolution, memes are the carriers of cultural evolution. And just as replication is a primary function of genes, replication (through imitation) is a primary function of memes.

2. Memes are complex ideas that are distinct memorable units—e.g. tunes, construction methods, catch-phrases, calculus, chess, postmodernism, evolution. (pp. 344-345)

3. Pictures, books, sayings are meme vehicles.

4. Some memes replicate themselves by functioning as filters screening out other memes seen as threatening, e.g. faith.

5. Differences between genes and memes that count against a science of memetics corresponding to the science of genetics.

a. There does not seem to be a meme counterpart to the DNA strands of genes.

b. Memes do not match the greater stability of genes (Note that this is not an absolute stability of genes--RY) because they tend to change so frequently with use.

c. There are significant commonalities however that make possible a scientific study of memes.

d. This scientific study is possible even if we presently lack details about the individual quirks of human psychology that affect memes.

6. Philosophical significance of memes

a. Memes, like genes, can be selfish replicators.

b. Memes, like genes, can survive because they are good replicators rather than just because of their fitness in their environments—although, over time, fitness is necessary for survival.

c. Memes are cranes that allow human beings (alone) to rise above, in some degree, the imperatives of our genes.

d. What we are as persons is a coalition of memes—with no one meme dominating and with no separate self standing apart from them.

C. Language, Meaning, and Mind

1. The Tower of Generate and Test - an oversimplified account of how the brain developed to enable organisms to find better and better ways of functioning

a. Darwinian creatures - functioning and then succeeding or failing through Natural Selection

b. Skinnerian creatures - blindly trying different actions and retaining those that are reinforced

c. Popperian creatures - having an inner environment with enough information about the outer environment to presort likelier successful moves prior to actually acting.

d. Gregorian creatures - being empowered by importing memes culturally (in particular, methods of proceeding) that allow superior functioning in the environment

2. Language (through culture) makes possible the joining of all human beings in one cognitive system for transmission of information in a way that far surpasses any capabilities of merely biological transmissions. (p. 381)

a. Anyone today understands ideas that would have been unthinkable to even geniuses of our grandparents' generation.

b. There is a link between language and intelligence that applies to human beings. Even when other creatures, such as chimps, can give evidence of many human-like characteristics, they still cannot perform some relatively simple human tasks such as tending a fire.

c. Chomsky was correct in describing language as being innate in the sense that children are naturally equipped to speak and understand languages; but he was incorrect in failing to recognize the evolutionary engineering processes (or cranes) that made this possible. (In effect, he attributed the origin of language to a skyhook.)

d. There is no reason why we cannot explain the origin of language through a gradual adaptionist design framework based on Natural Selection.

3. John Searle's insistence on "real intentionality," as opposed to as if intentionality, as a necessary property of minds in contrast with automatons is just an appeal to another skyhook.

a. As a critique of artificial intelligence, his insistence fails; and as a critique of evolutionary algorithmic development of the mind, it fails as well.

b. Searle is mistaken in claiming that real or original intentionality is necessary for an entity to have any real functions.

1) The fact that an artifact can exhibit functions not planned by the designer (e.g. the two-bitser machine being used for quarter-balboas) shows that function can be independent of real or original intentionality. (p. 407)

c. Although real intentionality has the advantage that it establishes special meanings by providing an originating causal role for the mind, Dennett sees no problem with discovering intentional meanings in a derived way from what actually occurs. (pp. 411-412)

1) His view supports development and interpretation of artificial intelligence in terms of function rather than real intentionality.

2) It also avoids what he sees as an appeal to a skyhook.

IV. Evolution and Ethics

A. Sociobiology

1. Bad Sociobiology

a. Bad sociobiology results from greedy ethical reductionism—e.g. E. O. Wilson's claims.

b. Bad sociobiology usually commits the genetic fallacy by presuming that because human behavior has genetic origins, current behavior must exemplify the genes in some simple way (pp. 472-473). There is a failure to recognize that (1) Cultural transmission has enormous power to change behavior (pp. 485-487); and (2) Functional adaptation can lead to current functions being different from ancestral function (p. 465).

1) E.g. a sudden strategy of protective defense can be just an excellent tool of rational thinking based upon current needs, desires, and limitations—rather than an exemplification of dormant defense genes.

2. Good Sociobiology

a. Any adequate ethics must take human nature into account. That is just Naturalism. And genetic predispositions are part of our human nature.

b. We should want to know about our biological limitations.

c. We need to remember however that human beings are "guided missiles" rather than projectiles on a ballistic track (p. 460)—and thereby are capable of making decision changes based upon our genes and memes.

B. Some Traditional Approaches to Ethics

1. Although religious texts may well lay out effective ethical codes or possibilities, they need to be subject to thorough testing rather than being accepted simply on faith (p. 476). In that sense, they are much like folk medicine: There may be some important truths there; but you had better proceed cautiously with thorough testing.

2. There is no algorithmic cost-benefit analysis based upon the Principle of Utility that can settle ethical questions because we cannot determine how to assign positive or negative values to outcomes (p. 498).

3. Similarly, Kant's Categorical Imperative is impractical as an algorithm because there are too many diverse ways of formulating the maxims to come under the rule of the Imperative (p. 500).

C. Rules - Dennett seems to accept some loose reference to rules in ethics—if only because ethics becomes more workable with rules and is not workable without them. But sometimes we also have to recognize that we are better off with fewer rules (p. 507)

D. A Set of Moral First Aid Manuals (pp. 508-509) - Rather than devising a rationally superior ethics, perhaps what we need is a set of different Moral First Aid Manuals (presumably based upon different traditional ethical approaches--RY) for different target audiences.

1. Each manual will strive constantly to improve the optimal practical design of a moral agent and to provide a context for interpretation of ethical situations.

2. Our own manual provides a "set of defaults" for purposes of screening ethical situations and we abide by the defaults unless attention-getting disruptions occur that demand more serious ethical rethinking of actions.

V. My Commentary

A. Affinities with Dennett - Generally, I strongly agree with Dennett's philosophical approach to problems through reasoning and science; and I agree with his naturalistic approach as well. But, as usual for these commentaries in the course, I will focus upon differences.

B. Memes - I find the concept of memes fascinating; but I do not think that they have the substance of genes. They appear to be an interesting, and perhaps very convenient, category of thought; but they have less empirical attachment to the external. (To use a term Dennett does not like, genes are natural kinds, whereas memes are not.) Memes are more like traffic lights and stop signs than the terrain on which vehicles travel. That is, the terrain is given (even though you can change it, of course, using physical methods of construction); but traffic lights and stop signs are one of many sets of conventional rules one can use to control traffic.

C. Artificial Intelligence - Since I have far less interest in Artificial Intelligence, I find too much of his philosophical interpretation skewed by his personal interests in AI. This does not mean that his hopes for advancements in AI will not be realized. It is just that I find him going to unnecessary lengths to show that AI is potentially equivalent to human intelligence.

D. Darwinism as a Universal Acid - Holding him to his universal acid claims, I would expect him to discuss emergence (emergent properties), since this concept fits so well with an evolutionary model and represents a departure from traditional notions. Yet he ignores emergence.

1. Emergence is the notion that a newly developed complex whole can be greater than the sum of its parts. What Dennett writes about evolutionary design, whereby ever more complex design occurs through a lengthy period of R & D (without intelligent direction), fits quite well with the notion of greater design complexity bringing about emergent properties.

2. Emergent properties are not a skyhook because they are the result of evolutionary processes, not the initiators of them; and there is nothing miraculous about them.

E. Mind and Consciousness - The concept of emergence makes possible recognition of mind, or consciousness, as an added dimension of reality that is dependent upon the presence of physical and organic processes, but is not entirely reducible to them.

1. The direct awareness of consciousness (a redundancy) is a sufficiently universal experience to warrant a strong claim as a dimension of reality. Likewise, the perceived (conceived) relevance of explanation of human actions in terms of consciousness (through reactions, intentions, motivations, decisions) is strong enough to require especially strong contrary evidence to dispel the perception.

2. The concept of emergence provides a justifying basis (a crane) for the existence and role of consciousness—minus the traditional baggage of human beings being created in the image and likeness of God, of the existence of a soul, of the mind as an entirely separate reality associated with a body, or of the mind as a simply autonomous free will.

3. Emergent consciousness provides the basis for a more robust conception of the self that is extremely advantageous in the development of ethics (quite aside from the likely personal satisfaction that it can produce).

F. Naturalistic Ethics - I would offer a much stronger naturalistic ethics than Dennett admits. Improving the ethical judgements and actions of human beings requires a diversity of approaches. But I want to concentrate here on just one element, namely, the relationship of facts to values and the significance of fundamental principles.

1. Dennett correctly points out difficulties in a cost-benefit calculation based upon the Principle of Utility. (In defense of Mill, I might point out that Mill was aware of these difficulties but hoped that increased experience would continually improve our calculations, a reasonable hope I think, and maintained that the Principle, regardless of difficulties, was superior to other approaches to ethical decision-making.) But I think that Dennett overlooks the potential positive significance of such principles.

2. My own fundamental principle of morality is the PSR Principle: You should act so as to maximize the totality of power, satisfaction, and reality.

a. If we were to approach significant ethical situations in life with an initial algorithmic cost-benefit analysis in terms of the PSR Principle, we would have a difficult time establishing high probabilities for courses of action, although I think that we could get a good start with this sort of analysis under some conditions.

b. More importantly though, focusing on the PSR Principle shows how one can tie values to facts—namely, by showing that you cannot function typically as a human being without valuing power, satisfaction, and reality. Establishing an affirmative relationship between facts and values (as opposed to a merely negating relationship where facts veto some values) is important for the development of naturalistic ethics.

c. The PSR Principle also provides a powerful R & D tool for analyzing actions already taken by a person in order to improve future judgments and actions. That is to say, we can gain a lot of insight into a past action by examining how the actually present perceptions of power, satisfaction, and reality led to the action. And if we choose to disagree with the action, we can initiate a discussion of whether or not those perceptions really maximized the totality of power, satisfaction, and reality.

Return to Lecture Notes Page

Return to Home Page

E-mail Messages:? ronald.yezzi@mnsu.edu

Last updated 12/12/01