The subject of this blog post might as well be catalogued as being among those things that scientists say that makes my head explode. In this case, sitting in the Bloomington Starbucks across from Sample Gates about a month ago, I heard a cognitive science (currently dissertating) PhD candidate say something to the effect: “It’s raw data, so there is no possibility of it being biased.” He was talking to a colleague, defending against some onslaught presented by a journal article, the title of which I didn’t catch. What I want to emphasize is the erroneous thinking of this student, who has since this time successfully defended his PhD thesis. I shake my head at this kind of lack of understanding so many scientists have of their own field and the general nature of science. Particularly egregious was his follow-up comments, which asserted that biasing cannot be added to unbiased data without it being extreme and obvious to all, as if the heavens would open and Zeus would callout, “biased!,” if such were to happen. I’ll only deal with the first statement that I paraphrased above.
I find it remarkable that so few scientists have heard of this notion, called “theory-ladenness.” For this reason, I think that the project of such institutions as Lyman Briggs College are very, very valuable to the arena of science, because their project’s aim is to educate scientists in way that their general education requirements are largely predicated on history of, philosophy of, and sociology of science —which is to say, they are trained to understand science just as well as they are trained to do science. These few scientists would understand that theory-ladenness means that instruments that produce data contain in them structure that is sensibly understood only through possession of that theory. It is like cryptography: someone with the instrument, who goes about producing data, does not know what the data means unless that person has a key (i.e., the theory). Having had this conceptual framework imposed upon the material object, the instrument, the object and its data are inherently biased. There is a reason we call the electrical current baseline in a discriminator of any digital device a “bias,” namely, because there is an arbitrary bias imposed by the scientist, to begin with. Anyone wishing to learn more about these ideas, I highly recommend looking to Norwood Russell Hanson’s work, as well as Pierre Duhem’s.
The above is bound to cause a reaction among readers not so knowledgeable about the more classical history of philosophy. What about the “raw” data of phenomenal experience? Actually, this is why Kant’s Critique of Pure Reason is so important. To summarize what that book talks about by asking a single question I posed to the cognitive science reading group I occasionally attend: How does the mind first cognize something without already knowing what it is that is to be cognized? This is a difficult question, and much of cognitive science (neuroscience, psychology, etc.) has put forward many efforts to answering this question. Even philosophers are still struggling with this sort of question, whether from a contemporary standpoint or a more classical philosophical standpoint (e.g., Adrian Johnston’s “The Genesis of the Transcendent: Kant, Schelling, and the Ground of Experience”). Kant, for example, thought that the mind was already availed of categories, from which experience could be constructed. It seems like a good idea; there are problems (see Johnston’s article tracing the elusive ground of experience). In short, it seems like the world doesn’t give to the mind “raw” data, but who knows. Philosophy and science are far from an initial decision on this, or, if they are, they lean toward the sentiment that there probably isn’t a such thing as raw data of experience.