If you have done any butchery or even carved the meat at the table, you will understand this metaphor. In order not to hack and end up with a terrible mess, you must follow the actual anatomy of the meat. In particular, the place to separate two bones leaving their muscles attached is at the joint. That is where you cut and break the two bones apart. This was Plato’s metaphor for making valid categories, ones that fit with the underlying ‘anatomy’ of nature.
It seems to me that we are not cutting at the joint in neuroscience. How does a science know if its concepts, categories, technical terms, contrasts/opposites are mirroring nature? Well, strictly speaking, there is no way to know that our categories are in keeping with nature. However, we can tell ways in which they are not. Perfection may not be possible but improvement almost always is. When we have to make room for odd little exceptions, when we can’t use the categories to make good predictions, when they are not easy to use, when they seem fragile to cultural or semantic differences, when they seem part of a slippery slope, when they do not fit with our theories – we have to think again about where the joints might be.
Why should neuroscience be in trouble with its categories? First, it is a very new science. It only really started in the last century; some would say it didn’t get going until into the 1980s and that would be 30 years ago. It does not have any overarching theory (not like Relativity, Quantum mechanics, Molecular theory, Plate Techonics, Cell theory, Evolution and the like). Its territory is more in ignorance than in light. Finding the joints is almost a matter of luck.
Second, it is immensely complex.
Third, neuroscience has inherited a lot of folk psychology; a great burden of Freudian psychology and other older theories; medical terminology and theories to do with mental illness; dated biological theories; attempts to simulate thought with computers; philosophical, legal and religious notions and theories. It is little wonder that agreed categories are next to impossible at the present time.
Take schizophrenia as an example. Most people treat that name as denoting a single disease. But it is more likely to be denoting a variety of diseases with differing causes, courses, symptoms, treatments, outcomes. There is no reason to accept, and many reasons to doubt, that it is a single disease. So what exactly does a statement like, “people suffering from schizophrenia hear voices”, mean. Not all schizophrenics hear voices and not everyone who hears voices is schizophrenic. And so it is with most symptoms of this ‘disease’. The same problem dogs ‘autism’ and some other conditions.
Intelligence is also hard to see as a clean category. How can it be measured? Is it one general thing or many specific one? Which specific ones? Do we know what personality is? Can we agree on subdividing it? What is its relationship to other things? There are so many, many words with such vague meanings. Neuroscience has words acquired from many sources. I read a philosophical paper and I wonder where do these words touch physical reality? What, I wonder, is a ‘mental state’; could it be a real thing? The popular press and some academics talk of ‘ego’. That is a Freudian concept and his division of the brain (ego, superego, id) is very clearly not at any ‘joints’? The computer set uses ‘algorithm’; just where are we likely to find algorithms in the brain?
It would seem that the closer a scientist is working to the level of cells and cell assemblies, the more likely they are to see the joints. But they would be less likely to be answering questions that people outside of neuroscience want answered. But unless people want to wade through oceans of muddy water, they may have to wait for answers to ‘important’ questions until after many boring questions have been investigated. My guess would be that the semantic arguments will continue because the words in which people are thinking are not doing a good job of the carving.