Cognitive Science is constituted by the following disciplines:
 

In principle, an indefinite number of disciplines can contribute to an interdisciplinary study of the mind.  These are the fields most commonly included.  And depending on your perspective, the field can be defined differently.  For example,  Cognitive Neuroscience is composed of two fields, cognitive science and neuroscience, though in the above listing neuroscience is just one discipline in a laundry-list of disciplines within cognitive science.

Cognitive science studies the mind at different levels of detail:  people have minds, which are seated (more or less) in their brains and central nervous systems.  Those brains are composed of neurons, which can (in principle) be described in biochemical terms, etc. To some extent, then, we can view the above disciplines as contributing to description of mental phenomena at different levels of detail:  for example, we could describe the physiological events that correspond to listening to this lecture at the level of all the biochemical interactions between brain cells, or by talking about the global architecture of the brain.  In the latter case, we would talk about the interaction between auditory and visual regions of the cortex, whereas in the former case these terms would not appear, being replaced by talk of individual cells and their relations.

(For the study of religion, we often have to talk about beliefs that are shared among a group of people, or practices that involve the cooperation of many people --- in other words, we deal with social phenomena.  There is a psychological component to social interactions, but the details between the social and the psychological are fuzzy at best, though in principle the disctincion seems quite clear.)

Caveats: 

In principle, then, we should be able to reduce more complex (higher-level) phenomena to simpler (lower-level) ones.  If we can, the higher levels can be explained by the lower levels, along with rules of intertheoretic reduction.  So we can get an overall picture as follows:

(If you have trouble printing the above graphic, click here and try printing again.)

Caveats regarding intertheoretic reduction: 

Now, reduction can be interpreted in several ways.  We want to say we can reduce one theory to another, but it isn't immediately clear what this means.  We can try to rephrase all the laws of one theory in different terms; here we are replacing all the predicates of a higher-level theory with predicates of a lower-level theory.  Or we can show small-scale models "fit into" large scale models; this gives us reason to think that reduction of laws is possible, even if technical problems remain in the details.  Again, once we reduce down to lower levels, we can then explain large-scale phenomena by demonstrating that they reduce to small-scale phenomena.

For example, we can understand how atoms combine into molecules by looking at atomic structure:  changing the orbits of electons either requires added energy or gives off energy, and thus only certain combinations of atoms form stable units.  But being able to give an account in terms of models of atoms and models of molecules doesn't mean we can state all chemical laws in terms of physics.

A different kind of explanation involves functions.  We can look at how parts of a system relate to one another in an abstract way, without resorting to technical details of the instantiation of these functions in particular objects.  We probably won't be able to achieve the same sort of reduction in these cases as before (though functional models can be decomposed into component parts, this isn't quite the same thing), but that is alright.  Different materials can be used to create the same functional states, so you don't  always want reduction all the way down from functional descriptions to physical descriptions.  The power of functional explanations is that the physical instantiation might be different in different cases.

So the relation of the different disciplines within cognitive science is not simple.  For example, a computer can be used to model the connectionist architecture of the brain; in this case the goal is to model the causal relations that account for mental activity.  A key to this application is the fact that the computer accurately models underlying physical processes. A different computer program could  be used to model the spread of an idea through a population.  In this case it isn't clear what physical processes underlie the process.

The Moral.  To some extent, though, some disciplines will be concerned with higher levels of description, and others with lower levels.  In any case, the idea of higher and lower levels is useful; this course will focus on very high levels of abstraction from the underlying physiological processes.  How any of our discussion will be of use to people who are primarily interested in the biology of the mind is not clear at the outset.  Some facts about psychology are best addressed by studying physiology directly, others not.  But there are connections, even if they aren't clear to us.  Andrew Newberg, an Assistant Professor of Radiology at Penn, has been in the news recently for work he has done on the physiological processes underlying meditation.  What he has demonstrated is not entirely clear, but he has at least shown how the same phenomenon can be described consistently at multiple levels.
 

The Biology of the Brain (a rudimentary lecture)

The basic component of brains is the neuron.  These are highly specialized cells that allow brains to be the seat of minds.  There are billions of cells in the brain, and each neuron makes contact with about 10,000 other neurons.  Because there are so many connections between so many elements, electrochemical signals can be passed through networks of neurons very quickly.  Dendrites carry small electric currents to the body of the neuron (I don't have a picture of this to post on the web, but will show a picture in class).  When the net contribution of individual currents reaches a certain amount, the cell will discharge this stored electric charge along the axon.  The charge will then be carried on to other neurons.  The physical mechanisms here are quite complex, and there is no need to consider them here.  But the result is that networks of neurons (neural nets) operate in a particular fashion.  Next class, when we discuss symbolic and connectionist architectures, understanding neurology at this basic  level of detail will be important.

The brain is composed of large, macroscopic parts.  People have assumed for many centuries that different parts of the brain do different things, but only in the last 150 years or so have we had good evidence for what parts perform what functions.  There are two caveats:  identifying the parts and identifying the functions.  For the former, click here.  Wernicke's area is involved in the processing of speech, and Broca's area in its production.  The latter has mostly been done through the study of pathologies:  people whose brains don't work the way most brains do, either because of accident or because of disease.  More recently, direct stimulation of the brain has been possible.
 

Mindware as Software

Mind and brain are conceptually different, even if the brain is the seat of the mind.

Interesting facts about minds:

Problems that the study of mind poses for sciences:  how can causal laws be efficacious without violating conservation laws?  (This is the mind-body problem in philosophy.)

Computation.  Consider formal logic, for example the following rule (modus ponens):

Formal rules are operations on pure tokens, i.e., symbols with the meaning removed.  We can substitute "Socrates is a philosopher" for p above and "Socrates is human" for q.  But we could also substitute another values.  p and q are tokens, and can be assigned any value.

Another example:  chess.  Chess is a formal system.  It can be played with any pieces so long as there are a certain number of pieces of a certain number of types (e.g., one king and two bishops).  The pieces can even be symbols on a computer screen.  Of course, to play chess one has to follow the rules of chess.

A computer as a physical device (rather than an abstract device) is a machine that will follow formal rules all on its own.  Given an imput, it will automatically execute processes and produce an output.  The physical properties of the machine matter in the sense that it is because of those properties that a machine can serve as a computer.  But otherwise they don't matter.

One of the key elements of cognitive science is the idea that brains operate like computers, in other words they are computers.  But extensive research is necessary to learn how precisely they function.  This can only be done indirectly, since it isn't possible to cut a brain open and watch everything that goes on inside it.  So neurologists observe bits and pieces of the behavior of brains and brain parts, and try to build a complete picture of how brains do what they do.  And computer scientisits build computers built of things other than brain parts that perform the same functions as brains.  This gives us two avenues from which to describe the formal properties of brains.  In other words, we can describe the physical properties that allow brains to be computers, and we can describe the algorithms that brains follow in carrying out tasks, but we can't actually watch brains carry out those tasks.
 

Two Topics

In the readings for next class are brief discussions of mental representations and cognitive architecture.  These outline the particular ways that the previous discussion of neurology and computation is relevant to human thought processes.
 

Cognition and Culture

To get an idea of the field that is taking shape around the issues that we will study in this course, consider the following mission statment for the Journal of Cognition and Culture (executive editors are E. Thomas Lawson and Pascal Boyer, vol. 1 published in 2001):
 

The key to this mission statement is the characterization of the subject matter of the journal:  "explanations of cultural phenomena in terms of acquisition, representation, and transmission involving common cognitive regularities without excluding the study of cultural differences".  This is more general than what we will study in this course, because it considers all sorts of cultural representations and not just religious ones.  The issues published so far are also narrower, in that they focus on the acquisition and transmission of representations; this is in keeping with research in the field, which is currently being supported by a handful of people.

Throughout the course we should keep in mind this charaterization of the field, and we should consider the following questions: