Wilfrid Sellars: Philosophy and the Scientific Image of Man

Section I. The Philosophical Quest

For Sellars, the aim of philosophy is a special sort of know-how: knowing how it “all hangs together.”  As such, philosophy isn’t a distinct subject matter or special science unto itself. Philosophy is distinct because it always has its eyes on the whole.  And as Sellars will aim to show, having such a “synoptic vision” is an achievement, for there are (at least) two seemingly conflicting perspectives of man in the world – a manifest image and a scientific image – that seemingly resist being fused together into a coherent whole.

Section II.  The Manifest Image

The manifest image is the conceptual framework in terms of which humanity first encounters itself as such, and so in a sense, it’s the image by which humanity first becomes the distinctive sort of discursive being that it is.  The manifest image is that by which we seek to determine the apparent correlations that we observe, without trying to explain those correlations in terms of the theoretical postulates posited by scientific thinking.  The manifest image is initially populated with persons and the resources to describe their characteristic actions.  Manifestly, we are in a world that includes other persons with distinctive characters, with whom we share obligations.  This image has been subsequently refined to include other ordinary objects and their manifest properties, and it may indeed be profitably refined further still to capture the world of sophisticated common sense.

Section III.  Classical Philosophy and the Manifest Image

The task of the “perennial philosophy” is to articulate the contours of this manifest image, in particular our abilities to engage in conceptual thought.  Such a task is objective, in that the manifest image may be more or less accurately described.  One of the shortcomings of pre-Hegelian philosophy is its lack of emphasis on the role of the group in shaping our understanding of the intelligible order by providing intersubjective standards by which our conceptual activities may be evaluated.  When we start to ask the how question: just how do we concept-mongering creatures come to have an image of ourselves in the world, we begin to engage in properly scientific thinking, a kind of thinking which will eventually threaten the entire framework around which the manifest image is constructed.

Section IV.  The Scientific Image

The scientific image, by contrast, is not concerned with establishing manifest correlations, but rather with explaining those observed correlations by postulating theoretical entities.  [Think here of the germ theory of disease, which postulates that living bodies may become colonized by unseen organisms, which rampantly reproduce themselves and overwhelm the host environment.]  The scientific image is actually composed of several nesting levels of such explanations, the higher-order postulates of which may be thought of as identical to complexes of lower-order postulated entities.  To show that different theories are so unified, however, does not mean that the higher order science may be replaced by the lower-order one, since the instruments and techniques of one science need not be reducible to that of the other.   

Behavioristic psychology, insofar as it eschews postulating theoretical entities altogether, properly belongs to the manifest image.  However, to the extent that psychology posits interior goings-on inside the brain of an organism (or thoughts), which systematically change according to changes in the dispositions of an organism to what it is disposed to say (or maybe do), then it belongs to the scientific image.  Such a postulated inner language of thought probably doesn’t do much to illuminate earthworm behavior.

Section V.  The Clash of the Images. 

So what happens when these two images conflict?  Which image should we take as describing reality itself?  The issue is pressing, since the scientific image identifies persons and other ordinary things as complex systems of imperceptible particles.  However, it is hard to identify a complex system of imperceptible particles with, say, as a pink ice cube, for how could we ever really understand such a complex to be pink?  There is an apparent “explanatory gap” between the structure and dynamics of the complex of imperceptibles and the property of being pink (why, for instance, wouldn’t it be lime green instead).  The traditional empiricist response to this difficulty is to locate the pinkness of the ice cube inside the mind of the observer as an “appearance,” not part of external reality itself.  But to do this, in turn, seems to lead one inexorably to a dualistic conception of the human person, since it is equally hard to see how a complex of imperceptible properties in the brain could itself be identified with a pink sensation.  [P. 16: This is perhaps an expression of the same overall dualistic inclinations that drive Chalmers’ conceivability argument.]  It might seem, then, that we should think of the postulates of the scientific image as ultimately unreal.

Section VI.  The Primacy of the Scientific Image: A Prolegomenon

But Sellars is a scientific realist; he thinks that the entities postulated by the scientific image are what really exist.

Conceptual thinking doesn’t present an insurmountable hurdle here, because in introspection, thoughts do not present themselves with any intrinsic qualitative character.  [Conceptual thought can be functionalized.]  Sensation and feeling, however, presents the greater challenge.  For such feelings are not characterized by their roles, but rather by their intrinsic characters.  And sensations themselves seem to possess an ultimate homogeneity that is hard to see could be reconstructed out of a complex system of imperceptible particles.  [I’m still having trouble with this section!]

Section VII.  Putting Man Into the Scientific Image

There is a further question about how to fit persons into the scientific image.  And Sellars thinks that in principle, a complete reconstruction of the notion of a person is impossible.  The reason is that the notion of a person is to some degree a normative notion.  One does more than simply describe a person when one attributes a personal predicate to them; one also conceives of them as belonging to a community and so bound up in a network of rights and responsibilities.  And to think thoughts of what a creature ought or ought not do is not to describe or to explain, but rather to rehearse a common, collective intention, without which meaningful discourse and rationality would be impossible.  The way, then, to complete the scientific image is to enrich it with the normative language of community and individual intentions (or the Brandomian language of commitment and entitlement).

Whew!

Some musings:

It isn’t as if Sellars rejects the manifest image and perennial philosophy altogether. Quite the opposite; he suggests that the scientific image needs to be supplemented with certain normative concepts that find their natural home in the manifest image. Instead, he might just be suggesting that the manifest image is not the measure of ultimate reality. Consequently, we might be free then to justify different aspects of it on pragmatic grounds.

Think again of games, chess pieces might not be items in “ultimate reality,” but that doesn’t preclude us from occasionally characterizing a certain blob of plastic as a pawn. Doing so does more than just describe it; it places the blob within a network of rules specifying what ought or ought not be done with it. Note again that our abilities to play something like chess with things like blobs of plastic still requires some sort of cooperation by the world. Not anything can profitably be regarded or used as a chess piece.

Perhaps characterizing a brain state as a sensation or a body as a person is similar. Although persons and sensations might not be the occupants of ultimate reality (reality as science takes it to be), that does not mean that we cannot profitably characterize certain complex physical goings-on as sensations or persons. Doing so does more than just describe those states; it also situates them in a normative framework characteristic of the realm of interpretation.

If all this is correct, then it may be a mistake to think of sensations and the like as having any sort of distinctive intrinsic or “real” nature. Saying that we are manifestly conscious though not really so might provide us with a new to say that we aren’t “really” conscious without having to give up the language of sensation and “what it’s like.”