Hack1.Find Out How the Brain Works Without Looking Inside

Hack 1. Find Out How the Brain Works Without Looking Inside

How do you tell what's inside a black box without looking in it? This is the challenge the mind presents to cognitive psychology.

Cognitive psychology is the psychology of the basic mental processesthings like perception, attention, memory, language, decision-making. It asks the question, "What are the fundamental operations on which mind is based?"

The problem is, although you can measure what goes into someone's head (the input) and measure roughly what they do (the output), this doesn't tell you anything about what goes on in between. It's a black box, a classic reverse engineering problem.¹ How can we figure out how it works without looking at the code?

These days, of course, we can use neuroimaging (like EEG [Hack 2], PET [Hack #3], and fMRI [Hack #4]) to look inside the head at the brain, or use information on anatomy and information from brain-damaged individuals [Hack #6] to inform how we think the brain runs the algorithms that make up the mind. But this kind of work hasn't always been possible, and it's never been easy or cheap. Experimental psychologists have spent more than a hundred years refining methods for getting insight into how the mind works without messing with the insides, and these days we call this cognitive psychology.

There's an example of a cognitive psychology-style solution in another book from the hacks series, Google Hacks (http://www.oreilly.com/catalog/googlehks). Google obviously doesn't give access to the algorithms that run its searches, so the authors of Google Hacks, Tara Calishain and Rael Dornfest, were forced to do a little experimentation to try and work it out. Obviously, if you put in two words, Google returns pages that feature both words. But does the order matter? Here's an experiment. Search Google for "reverse engineering" and then search for "engineering reverse." The results are different; in fact, they are sometimes different even when searching for words that aren't normally taken together as some form of phrase. So we might conclude that order does make a difference; in some way, the Google search algorithm takes into account the order. If you try to whittle a search down to the right terms, something that returned only a couple of hits, perhaps over time you could figure out more exactly how the order mattered.

This is basically what cognitive psychology tries to do, reverse engineering the basic functions of the mind by manipulating the inputs and looking at the results. The inputs are often highly restricted situations in which people are asked to make judgments or responses in different kinds of situations. How many words from the list you learned yesterday can you still remember? How many red dots are there? Press a key when you see an X appear on the screen. That sort of thing. The speed at which they respond, the number of errors, or the patterns of recall or success tell us something about the information our cognitive processes use, and how they use it.

A few things make reverse engineering the brain harder than reverse engineering software, however.

Biological systems are often complex, sometimes even chaotic (in the technical sense). This means that there isn't necessarily a one-to-one correspondence in how a change in input affects output. In a logic-based or linear system, we can clearly see causes and effects. The mind, however, doesn't have this orderly mapping. Small things have big effects and sometime big changes in circumstance can produce little obvious difference in how we respond. Biological functionsincluding cognitionare often supported by multiple processes. This means they are robust to changes in just one supporting process, but it also means that they don't always respond how you would have thought when you try and influence them.

People also aren't consistent in the same way software or machines usually are. Two sources of variability are noise and learning. We don't automatically respond in the same way to the same stimulus every time. This sometimes happens for no apparent reason, and we call this randomness noise. But sometimes our responses change for a reason, not because of noise, and that's because the very act of responding first time around creates feedback that informs our response pattern for the next time (for example, when you get a new bike, you're cautious with your stopping distance at first, but each time you have to stop suddenly, you're better informed about how to handle the braking next time around). Almost all actions affect future processing, so psychologists make sure that if they are testing someone the test subject has either done the thing in question many times before, and hence stopped changing his response to it, or he has never done it before.

Another problem with trying to guess how the mind works is that you can't trust people when they offer their opinion on why they did something or how they did it. At the beginning of the twentieth century, psychology relied heavily on introspection and the confusion generated led to the movement that dominated psychology until the '70s: behaviorism. Behaviorism insisted that we treat only what we can reliably measure as part of psychology and excluded all reference to internal structures. In effect we were to pretend that psychology was just the study of how stimuli were linked to outputs. This made psychology much more rigorous experimentally (although some would argue less interesting). Psychology today recognizes the need to posit mind as more than simple stimulus-response matching, although cognitive psychologists retain the behaviorists' wariness of introspection. For cognitive psychologists, why you think you did something is just another bit of data, no more privileged than anything else they've measured, and no more likely to be right.²

Cognitive psychology takes us a long way. Many phenomena discovered by cognitive and experimental psychology are covered in this bookthings like the attentional blink [Hack #39] and state-dependent recall [Hack #87] . The rigor and precision of the methods developed by cognitive psychology are still vital, but now they can be used in tandem with methods that give insight into the underlying brain structure and processes that are supporting the phenomenon being investigated.

1.2.1. End Notes

1. Daniel Dennett has written a brief essay called "Cognitive Science as Reverse Engineering" (http://pp.kpnet.fi/seirioa/cdenn/cogscirv.htm) in which he discusses the philosophy of this approach to mind.

2. A psychologist called Daryl Bem formalized this in "self-perception theory." He said "Individuals come to know their own attitudes, emotions and internal states by inferring them from observations of their own behavior and circumstances in which they occur. When internal cues are weak, ambiguous, or uninterpretable, the individual is in the same position as the outside observer." Bem, D. J., "Self Perception Theory." In L. Berkowitz (ed.), Advances in Experimental Social Psychology, volume 6 (1972).