Read Darwin's Dangerous Idea Online
Authors: Daniel C. Dennett
build dams that take months to complete, couldn't chimpanzees tend a simple campfire? This rhetorical question climbs a misleading ladder of abilities. It ignores the independently well-evidenced possibility that there are two profoundly different ways of building dams: the way beavers do and 2. This is an elaboration of ideas I first presented in Dennett 1975.1 recently discovered that Konrad Lorenz (1973) described a similar cascade of cranes—in different terms, of course.
374 LOSING OUR MINDS TO DARWIN
The Role of Language in Intelligence
375
In the beginning—once the pump had been primed—there was Darwinian evolution of species by natural selection. A variety of candidate organisms were blindly generated by more or less arbitrary processes of recombination and mutation of genes. These organisms were field-tested, and only the best designs survived. This is the ground floor of the Tower. Let us call its inhabitants
Darwinian creatures.
Skinnerian creature "blindly" .. . until one is selected by Next time, the creature's first tries different responses . . .
"reinforcement."
choice will be the reinforced response.
FIGURE 13.2
options before risking them in the harsh world. We human beings are creatures capable of this third refinement, but we are not alone. We may call the beneficiaries of this third story in the Tower
Popperian creatures,
since, Darwinian creatures, different
selection of one favored
multiplication of the
as Sir Karl Popper once elegantly put it, this design enhancement "permits
"hard-wired" phenotypes
phenotype
favored genotype
our hypotheses to die in our stead." Unlike the merely Skinnerian creatures, FIGURE 131
many of whom survive only because they make lucky first moves, Popperian This process went through many millions of cycles, producing many creatures survive because they're smart enough to make better-than-chance wonderful designs, both plant and animal, and eventually among its novel first moves. Of course, they're just lucky to be smart, but that's better than creations were some designs with the property of phenotypic plasticity. The just being lucky.
individual candidate organisms were not wholly designed at birth, or, in other words, there were elements of their design that could be adjusted by events that occurred during the field tests. (This is what makes the Baldwin Effect possible, as we saw in chapter 3, but now we are going to focus on the intra-organismic design that sets up that crane.) Some of these candidates, we may suppose, were no better off than their hard-wired cousins, since they had no way of favoring ( selecting for an encore ) the behavioral options they were equipped to "try out," but others, we may suppose, were fortunate enough to have wired-in "reinforcers" that happened to favor Smart Moves, actions that were better for their agents. These individuals thus confronted the Popperian creature has an inner
First time, the creature acts in a
environment by generating a variety of actions, which they tried out, one by selective environment that previews
foresightful way (better than chance).
candidate acts.
one, until they found one that worked. We may call this subset of Darwinian creatures, the creatures with conditionable plasticity,
Skinnerian creatures,
FIGURE 13-3
since, as B. F. Skinner was fond of pointing out, operant conditioning is not But how is this preselection in Popperian agents to be done? Where is the just analogous to Darwinian natural selection; it is continuous with it. "Where feedback to come from? It must come from a sort of
inner environment
—an inherited behavior leaves off, the inherited modifiability of the process of inner something-or-other that is structured in such a way that the surrogate conditioning takes over" (Skinner 1953, p. 83).
actions it favors are more often than not the very actions the real world would Skinnerian conditioning is a fine capacity to have, so long as you are not also bless, if they were actually performed. In short, the inner environment, killed by one of your early errors. A better system involves
preselection
whatever it is, must contain lots of
information
about the outer environment among all the possible behaviors or actions, weeding out the truly stupid and its regularities. Nothing else ( except magic ) could provide preselection 376 LOSING OUR MINDS TO DARWIN
The Role of Language in Intelligence
377
worth having. Now, here we must be very careful not to think of this inner hard-wiring under a new layer of pre-emptive control ( a theme developed in environment as simply a replica of the outer world, with all its physical con-the work of the AI researcher Rodney Brooks [e.g., 1991] ). It is these higher tingencies reproduced. (In such a miraculous toy world, the little hot stove in levels of control that have the potential for vast increases in versatility. And your head would be hot enough actually to burn the little finger in your head it is at these levels in particular that we should look for the role of language that you placed on it!) The information about the world has to be there, but it (when it finally arrives on the scene), in turning
our
brains into virtuoso also has to be structured in such a way that there is a nonmiraculous preselectors.
explanation of how it got there, how it is maintained, and how it actually We engage in our share of rather mindless routine behavior, but our achieves the preselective effects that are its
raison d'etre.
important acts are often directed on the world with incredible cunning, Which animals are Popperian creatures, and which are merely Skinner-ian?
composing projects exquisitely designed under the influence of vast libraries Pigeons were Skinner's favorite experimental animals, and he and his of information about the world. The instinctual actions we share with other followers developed the technology of operant conditioning to a very so-species show the benefits derived from the harrowing explorations of our phisticated level, getting pigeons to exhibit quite bizarre and sophisticated ancestors. The imitative actions we share with some higher animals may learned behaviors. Notoriously, the Skinnerians never succeeded in proving show the benefits of information gathered not just by our ancestors, but also that pigeons were
not
Popperian creatures, and research on a host of different by our social groups over generations, transmitted nongenetically by a "traspecies, from octopuses to fish to mammals, strongly suggests that if there are dition" of imitation. But our more deliberatively planned acts show the any purely Skinnerian creatures, capable only of blind trial-and-error benefits of information gathered and transmitted by our conspecifics in every learning, they are to be found among the simple invertebrates. The sea slug culture, including, moreover, items of information that no single individual
Aplysia
has more or less replaced the pigeon as the focus of attention among has embodied or understood in any sense. And though some of this those who study the mechanisms of simple conditioning. ( Researchers information may be of rather ancient acquisition, much of it is brand-new.
unhesitatingly and uncontroversially rank species in terms of how intelligent When comparing the time scales of genetic and cultural evolution, it is useful they are. This involves no myopic endorsement of the Great Chain of Being, to bear in mind that we today—every one of us—can
easily
understand many no unwarranted assumptions about climbing the ladder of progress. It ideas that were simply unthinkable
by the geniuses
in our grandparents'
depends on objective measures of cognitive competence. The octopus, for generation!
instance, is stunningly smart, a fact that would not be available to surprise us The successors to mere Popperian creatures are those whose inner en-if there weren't ways of measuring intelligence that are independent of vironments are informed by the
designed
portions of the outer environment.
phylogenetic chauvinism.)
We may call this sub-sub-subset of Darwinian creatures
Gregorian creatures,
We do not differ from all other species in being Popperian creatures, then.
since the British psychologist Richard Gregory is to my mind the pre-eminent Far from it; mammals and birds and reptiles and fish all exhibit the capacity theorist of the role of information (or, more exactly, what Gregory calls to use information from their environments to presort their behavioral options Potential Intelligence) in the creation of Smart Moves (or what Gregory calls before striking out. We have now reached the story of the Tower on which I Kinetic Intelligence). Gregory observes that a pair of scissors, as a well-want to build. Once we get to Popperian creatures, creatures whose brains designed artifact, is not just a result of intelligence, but an endower of have the potential to be shaped into inner environments with preselective intelligence ( external Potential Intelligence ), in a very straight-forward and prowess, what happens next? How does new information about the outer intuitive sense: when you give someone a pair of scissors, you enhance his environment get incorporated into these brains? This is where
earlier
design potential to arrive more safely and swiftly at Smart Moves (Gregory 1981, decisions come back to haunt—to constrain— the designer. In particular, pp. 31 Iff).
choices that evolution has already made between need-to-know and Anthropologists have long recognized that the advent of tool use accom-commando-team now put major constraints on the options for design panied a major increase in intelligence. Chimpanzees in the wild fish for improvement. If a particular species' brain design has already gone down the termites with crudely prepared fishing sticks. This fact takes on further need-to-know path with regard to some control problem, only minor significance when we learn that not all chimpanzees have hit upon the trick; modifications (fine tuning, you might say) can be
readily
made to the existing in some chimpanzee "cultures," termites are a present but unexploited food structures, so the only hope of making a major revision of the internal source. That reminds us that tool use is a two-way sign of intelligence; not environment to account for new problems, new features of the external only does it
require
intelligence to recognize and maintain a tool (let alone environment that matter, is to
submerge
the old fabricate one), but tool use
confers
intelligence on those who are lucky 378 LOSING OUR MINDS TO DARWIN
The Role of Language in Intelligence
379
enough to be given the tool. The better designed the tool (the more infor-and hiding, and their kin enter a child's brain, they find homes already mation embedded in its fabrication), the more Potential Intelligence it partially built for them. Ray Jackendoff (1993) and other linguists have confers on its user. And among the pre-eminent tools, Gregory reminds us, identified fundamental structures of spatial representation—notably designed are what he calls "mind-tools": words.
to enhance the control of
locomotion
and
the placement
of movable things—
that underlie our intuitions about concepts like
beside, on, behind,
and their kin. Nicholas Humphrey (1976,1983,1986) has argued that there must be a genetic predisposition for adopting the intentional stance, and Alan Leslie (1992 ) and others have developed evidence for this, in the form of what he calls a "theory of mind module" designed to generate second-order beliefs (beliefs about the beliefs and other mental states of others). Some autistic children seem to be well described as suffering from the disabling of this module, for which they can occasionally make interesting compensatory adjustments. (For an overview, see Baron-Cohen 1995.) So the words (and hence memes) that take up residence in a brain, like so many earlier design novelties we have considered, enhance and shape pre-existing structures, rather than generating
entirely
new architectures (see Sperber [in press] for a Darwinian overview of this exaptation of genetically provided functions by Gregorian creature imports mind-tools
culturally transmitted functions). Though these newly redesigned functions from the (cultural) environment; these
improve both the generators and the
are not made from whole cloth, they do create an explosive new capacity to testers.
look ahead.
FIGURE 13.4
An internal model allows a system to look ahead to the future conse-Words and other mind-tools give a Gregorian creature an inner environ-quences of current actions, without actually committing itself to those ment that permits it to construct ever more subtle move-generators and actions. In particular, the system can avoid acts that would set it irretriev-move-testers. Skinnerian creatures ask themselves, "What do I do next?" and ably down some road to future disaster ("stepping off a cliff"). Less dra-haven't a clue how to answer until they have taken some hard knocks.
matically, but equally important, the model enables the agent to make Popperian creatures make a big advance by asking themselves, "What should current "stage-setting" moves that set up later moves that are obviously I think about next?" before they ask themselves, "What should I do next?"