There is no way to establish fully secured, neat protocol statements as starting points of the sciences. There is no tabula rasa. We are like sailors who have to rebuild their ship on the open sea, without ever being able to dismantle it in dry-dock and reconstruct it from its best components.
If Martin Heidegger brought epistemological holism to Continental Philosophy through the central metaphor of the hermeneutic circle, W. V. Quine did the same for Analytic Philosophy. Part of how he did that was by popularizing the idea of Neurath’s Boat.
The idea is this: we can only appraise the truth-value of some specific assertion in light of a background web of beliefs that provide the necessary context. Thus, it is like we are sailing on a boat which we can scrutinize—and even repair—one plank at a time, but not all at once.
A more straightforward way of saying all of this is that you can only really understand the partial truths available to you in the context of the whole truth. But the whole truth is not available to anyone, and so instead we have beliefs, theories, frameworks.
Neurath’s Boat is a neat metaphor, but I prefer one I encountered from Susan Haack (via Joseph Heath). The idea is that knowledge is like a crossword puzzle. What you fill out has implications for what the likely answers are for nearby slots. But when you figure out what goes into those nearby slots, if you have greater confidence about them, you may need to revise what you put in the initial slot. Thus, the crossword puzzle has this property of a web of answers that are interconnected in important ways.
If we stuck with the idea of a puzzle we could see all of and were merely filling entries out for, it would imply an incrementalism just like the one explicit in Neurath’s Boat. But that would be silly—we do not know what the whole truth looks like. If we did, we would have a great deal more certainty about an enormous number of areas that remain hotly contested.
No, instead it seems to me that the crossword puzzle metaphor works, but we only get pieces of it and attempt to guess at the shape of the rest. Our experience of partial truths is fundamentally projective—based on the parts we encounter we attempt to project a provisional outline of the whole truth, though this outline may itself be quite incomplete.
When Hans-Georg Gadamer argues that the hermeneutic experience, as well as dialectic, are fundamentally about the subject rather than authorial intent, I think he means something like this: guessing at authorial intent is like trying to guess what their projection of the whole truth is. Focusing on the subject, on the other hand, is simply attempting to work out what we think the correct projection of the whole truth is, using what we learn from what we read or the conversation we participate in.
One last observation: the whole truth is not value-neutral; not for humans. Elizabeth Anderson argues this very convincingly—responding, coincidentally, to Susan Haack. She uses the example of The Secret Relationship, a book which apparently uses some very selectively chosen facts in a misleading way. “Although many characters in mysteries lie, the most interesting characters deceive by telling the truth—but only part of it.”
How are we to assess the significance of the facts cited in The Secret Relationship? Taken in isolation, they suggest that Jews played a special or disproportionate role in the Atlantic slave system or that their participation was more intense than that of other ethnic and religious groups. But in the context of additional facts, such as those just cited, they show that Jewish participation in the slave system was minor in absolute terms and was no different in intensity from similarly situated ethnic and religious groups. The larger context exposes a serious bias or distortion in the way The Secret Relationship characterizes the significance of Jewish participation in the Atlantic slave system. The characterization is “partial” in the literal sense that it tells only part of the truth needed to assess the significance of the matters at hand. What matters for assessing significance, then, is not just that an account be true but that it in some sense represent the whole truth, that it be unbiased. Furthermore, the fact that an account is biased or distorted is a good reason to reject it, even if it contains only true statements. Haack’s premise (2) is therefore false: to justify acceptance of a theory one must defend its significance, not just its truth.
Our projection of the whole truth has political implications, and so too do the partial truths accessible to us, because of the whole truth that they imply. There is therefore an importantly ethical dimension in how we assemble those partials truths.