For those of you in the Colorado Springs area I'll be speaking tonight at the University of Colorado on WHY PEOPLE BELIEVE WEIRD THINGS from 6:00-7:30 at the University Center Room 124 A & B, and I'll be on KDOR radio with Richard Randall from 4:00-6:00pm, apparently the #1 rated talk show in Colorado Springs.
For those of you in the So. California area Dr. Susan Blackmore will be in town on her book tour, lecturing for the Skeptics Society Wednesday night, April 21, at 7:30 pm at Baxter Lecture Hall, on THE MEME MACHINE. Dr. Blackmore, best known to skeptics for her research into the paranormal, out-of-body, and near-death experiences (see the interview with her in Skeptic Vol. 6, #3), has shifted her research interests to something a little more grounded in reality (at least she hopes that it is!), and that is taking Richard Dawkins' concept of the meme (briefly discussed in THE SELFISH GENE) and expanding it into a full theoretical model of how ideas infect minds and culture, and how, quite possibly, meme selection takes pride of place alongside natural and sexual selection as a major force in human evolution. Blackmore was featured in a two-page spread in Time magazine, entitled "The Selfish Meme," written by Richard Dawkins and based on his foreword to Susan's new book. Blackmore's book is daring. She goes so far as to speculate on the role of memes in the expansion of the human brain, still one of the deep mysteries for cognitive scientists and evolutionary biologists.
(Actually this problem goes back to my old friend Alfred Russel Wallace--I call him that because my doctoral dissertation was on Wallace and the problem of mind and brain for him and Darwin. Wallace could not conceive of how or why natural selection would construct a brain double the size it really needs to be to survive in a natural environment--of what purpose does mathematics, music appreciation, and other abstract thought processes serve in survival? Virtually none, Wallace believed, therefore there must be a higher intelligence that directed the development of mind and big brain. Also, since Wallace believed in spiritualism and even attended seances, the "psychic" nature off mind could also not be explained by natural selection, yet another place for divine inspiration. Darwin, of course, was dismayed at all this, writing to Wallace: "I hope you have not murdered yours and my child"--natural selection.)
We published the first article on memes by Susan in Skeptic Vol. 5, #2, and a strong critique of memes in Vol. 6, #3 by James Polichak, a cognitive psychologist at SUNY, Stony Brook. Polichak outlines what he sees as four basic shortcomings to memetics (obviously this is not a critique of Blackmore's new book since it just came out last week, but the general principles might apply):
1. Memeticists have not done an adequate job of defining the meme, nor have they offered any examples of what a meme might be that withstand scrutiny.
2. Memeticists have failed to show that memes are necessary to understanding culture. As a consequence they are unable to show that models based on biological selection are inadequate.
3. By largely ignoring the principles and data concerning information processing from the social sciences, especially psychology, memeticists have argued for a highly inaccurate model of information transfer, and a highly limited model of the activity of the human brain.
4. Memeticists have offered inaccurate and circular claims about what kind of explanatory power is obtained by assuming the existence of memes.
Further, Polichak adds:
With regard to how information is transmitted with potential mutation and is subject to selective forces leading to differential survival, the writings of memeticists are about as vague as their attempts to define the meme. It is also not clear to what extent we can meaningfully discuss transmission of information (as opposed to reconstruction of information). Memeticists have also not done enough to differentiate memetic transmission of information from non-memetic transmission. It is known that humans can transmit information to each other that could not reasonably be considered memetic. For example Russell, Switz, and Thompson (1980) showed that human menstrual cycles become synchronized through olfactory cues. Presumably there is some variance in the degree to which people's menstrual cycles become synchronized, but we would probably not want to say that this variability is evidence for mutation and differential survival of any particular menstrual cycle. It is up to memeticists to demonstrate that the information that they deal with is different, and this will prove difficult. Cognitive psychologists have demonstrated that learning and remembering are sensitive to environmental and perceptual factors, which are not considered in memetic analyses, and that most human thought is not likely to be memetic. They have also shown evidence for the recall of information never transmitted. Memeticists must show that, after accounting for these pieces of evidence and the psychological theories based on them, there is some form of discrete information left over that is subject to mutation (not merely variability) and differential selection (not based on perception, attention, or mental reconstruction of experience). In other words, they must demonstrate that, contrary to current psychological models, all forms of information in the human brain are not like the information discussed above before they can develop meaningful predictions and models of memetic transmission.
Blackmore's (1997) analysis of why we talk so much also conflicts, like so much in memetics, with psychological theory and research. Blackmore presents an extremely competitive model of the development of human language and its current use we don't seem to care much what anyone else has to say because we're just waiting for our turn so that we can transmit our memes. There is a growing body of experimental evidence in psychology for a collaborative theory of language use (e.g., Clark, 1992). Numerous experiments have examined the ways that speakers work together to decide what to call ambiguous objects. Parallels to figuring out where to go to dinner, how to put together a bicycle, and so on, should be obvious. According to Clark's collaborative theory, language is used by people so that they can attain a reasonable degree of mutual understanding of their environments and intentions in order to interact effectively. Like Dunbar's (1993; 1996) analysis of language development, Clark's theory is based on the idea that language is an important way to coordinate activity among people and to effectively describe and manipulate each other and the environment. Blackmore's ideas about language use and development seem far more limited and far less likely.
Blackmore (1997) offers three similar analyses for why we are so nice to each other, why our brains are so big, and why we think we have a self. In each case, of course, the answer is memes. Throughout her analyses, Blackmore asks the reader to continually "imagine a world full of brains, and far more memes than can possibly find homes." This pattern of thought imagining all those memes struggling to survive in the limited human brain, she suggests, will allow us to answer the difficult questions. Blackmore does not, however, offer any evidence for why and how memes might actually be the cause of our thoughts, big brains, niceness, and so on. We are asked to take our excessively big brains as evidence for the existence of memes and are expected to accept memes as a reason for our big brains existing. Blackmore, and other memeticists, are essentially asserting that memes are out there, without evidence or even an adequate example, and without regard for the conflict with psychological models. They then expect us to assume the existence of memes and insert that term as an answer to life's mysteries.
I hope that the above critique has shown that memeticists have grossly overstated the power of a memetic approach to understanding information processing and culture. They have much work to do to convince the skeptical scientist of the value of the meme, much less its existence. Memeticists should start by looking at the data from the social sciences and the models developed from them. They need to show that they can account for the objections put forth in this paper based on those psychological models and on logical grounds. Memeticists need to more clearly define the kinds of information they are going to deal with, and show that existing models are flawed when it comes to understanding this kind of information. Then they must demonstrate that the memetic approach can succeed where biological or psychological approaches have failed. Nothing presented in the memetics literature thus far suggests that memeticists will be able to accomplish this. Ill-considered examples, ignorance of relevant experimental research, and exaggerated claims of explanatory power do not make for a convincing scientific theory.
Next posting: A summary of Jared Diamond's talk on HOW TO BE RICH AND SUCCESSFUL: LESSONS FROM HISTORY'S EXPERIMENTS, delievered last Sunday at Caltech for the Skeptics Society, and well received (although not without some rather challenging questions).