Objectivity

In his latest book Agazzi (2014) introduces an illuminating distinction between things and objects. So far, I mostly used the generic and encompassing term “en­tity” to refer to things, objects, events, processes etc.

“Scientific objects exist as abstract objects (i.e. as intellectual constructions) that encode certain properties, while not being purely abstract since they are exemplified (within certain margins of accuracy) by concretely existing objects” (Agazzi 2014, p. 104, n. 48)

But a thing cannot be reduced to an object, since a thing also has properties which do not belong to the object it instantiates. Moreover, the instantiation of an object by a thing need not be exact in the sense that a property of an object might only approximate a property of a thing. An object is thus constructed by adopting a certain point of view, some perspective on a thing. An object cannot encode all the properties of a thing. The predicative statements that attribute the properties of an object to a thing only convey a partial knowledge of the latter.

When we attempt to represent some observed targets, we similarly clip out some relevant aspects of those targets and connect them with some properties of our representors.

In what sense can my representations be objective? When users agree that a representor adequately or correctly represents its target in some respects, we get objectivity in the ‘weak’ sense of intersubjectivity (Agazzi 2014, 51ff.). But such agreement does not guarantee per se that there is something in the world that instantiates the relevant properties, i.e. the aspects selected when constructing the object of the representation. Intersubjective agreement fails to substantiate the claim that there is a thing “out there” possessing properties which conform to those encoded by the object of the representation. Such conformity is what Agazzi calls ‘strong’ objectivity. Surely, I concede that we can collectively be wrong in attributing properties to things. I thus endorse fallibilism.

Now, van Fraassen is quite aware of the importance of reaching strong objec­tivity. However, he subscribes to a representational conception of knowledge, albeit not in the classical, psychological, sense which took representations to be mental ideas. Then, he must propose a solution of the bridge problem, which he baptizes “the loss of reality objection” according to which our representors, our scientific models, do not hit on something external.

“How can an abstract entity, such as a mathematical structure, represent something that is not abstract, something in nature?” (van Fraassen 2008, 240)

The answer he gives is pragmatic and amounts to a “dissolution” of the loss of reality objection. When I claim that my representation is adequate to the phe­nomenon I also claim that my representation is adequate to the phenomenon as represented by me (van Fraassen 2008, 253-261). I cannot assert that my graph is adequate to the evolution of the deer population in Princeton as represented by me and at the same time deny that it represents such evolution in a correct way. If I do this, I fall into pragmatic inconsistency.

I’m happy to concede that the indexical ingredient in the activity of representing, which is performed by me or us, prevents from positioning myself at a God’s eye point of view and enjoying an overarching view of my graph and the deer popu­lation. Yet, such a difficulty arises only when knowing is identified with repre­senting. Such epistemological posture entails that the loss of reality objection must be overcome by resorting to a pragmatic move.

Moreover, even if we grant, as I do, that within the representational conception of knowledge, the loss of reality objection is pragmatically dissolved, doubts can be raised about our capacity of knowing real things by means of our representors, above all if we adopt what I will call a predicative conception of knowledge. According to the latter conception which I favor, knowing is attributing to things properties which they exemplify and to have good reasons for doing so. Such acts of predication are expressed in statements such as “At t₀, the deer population in Princeton is n₀”. Within the representational conception of knowledge, strong objectivity collapses into weak objectivity and the bridge problem vanishes or “dissolves” as van Fraassen aptly says. Since we can't ascend to a godlike pedestal, we are forever confined to the domain of intersubjectivity or weak objectivity. Yet, such confinement is a consequence of the erroneous philosophical thesis according to which to know is to represent.

Notice that adequacy in van Fraassen's sense is not to be confused with truth, and certainly not with truth in a correspondence sense. Adequacy for him amounts to weak objectivity, that is, the invariance of some aspects of targets with respect to the particular points of view or perspectives of the various users of the same representors. For van Fraassen a representation is adequate or objective if it cor­rectly represents its referent, in the sense that the usually agreed upon verification procedures have been implemented and consensus has been reached on their out­comes. Contrary to van Fraassen, I insist that the correctness of a representation rests on the truth of some predicative statements which attribute properties to things.

3