You move to accept questions as a distinct epistemic element, and argue this by attempting to show that questions exist, they are not theories, and they are not methods. However, is it not also possible that questions may be functions of both theories and methods simultaneously; that is, if a theories and methods are propositional sentences of the form T and M respectively, then why can we assume that questions are not propositional sentences of the form f(T,M)? In this case, they would not be theories, nor would they be methods, but they would not be distinct from other epistemic elements either. I see no reason why questions could not be syntactically represented as a function of theories and methods, and this is nowhere elaborated upon. Thus, I do not think your conclusions follow, as there is an addition case to be dealt with.
Moreover, since you did not formally treat the syntax of questions (you mentioned erotetic logic, but never expanded on it, and never presented its formalism), it is not entirely clear to me that, barring this first problem, you succeeded in demonstrating you claims. Also, take note that questions which are not why questions can always be answered by a theory (Q:"where is the University of Toronto?" A: "The university of Toronto is in Toronto." for instance, is a theory) and so it is not clear that why questions need be an exception to this. Furthermore, the answer to a why question could conceivably take the form of a method (given the question "why A?", the answer is "A because B", however, this may be read like equivalent to "if B, then A", which is has a higher-order structure identical to that of a method).
There also seem to be several instances where you brush over the formal details of the arguments you are making (as though you are saying they are trivial or so intuitive that they need no elaboration) but these details are not obvious and are very important to your argument (see for example the last sentence of paragraph 2 on page 4). These results may be trivial if you consider the formal construction of your arguments, but since this is not given, I cannot see why these may be so easily dismissed.
Finally, you talk about answering questions, and use the notion of logical implication as almost synonymous with answering (see the second last paragraph of page 8 for example) however, you never define what an answer is, and then proceed to use important details of such a definition in your argument. This is lacking because it seems as though you use the term in different contexts as it is convenient in that context.
I would move to accept questions as a distinct epistemic element if a more formal proof with explicit syntactic arguments, and if every instance where a logical result is taken to be trivial were instead explicated, even if only briefly, and if all possible cases are considered and the same conclusion is reached. If these actions are carried out and the conclusion that questions are distinct from theories and methods no longer holds, then I would move to reject this change. If such arguments cannot be provided, I would move to reject, on the basis that the claims made here lack rigour. I think the topic of this paper is quite interesting, and moreover, I think that it is very important to scientonomy, hence my push for more rigorous treatment. I am sorry that this comment is rambling. I can provide a more detailed explanation of these points, and why I think they are of crucial importance at the author's request.
I’d like to begin my response to Patrick Fraser's comment by calling attention to the distinction between two different tasks:
1. evaluating the formal structure of Will Rawleigh's argument for irreducibility of questions to theories/methods;
2. deciding whether the ontology of scientific change is better with or without questions as an epistemic element.
I agree with Patrick that the former task is extremely important. I do agree with Patrick that the formal proof for the irreducibility of questions to theories/methods is lacking. More research is needed to formally prove that questions are not reducible to theories and/or methods; I hope that scientonomists with a training in formal sciences will take on this important task going forward. Yet, our task here is different. Our current question is whether the ontology of scientific change with questions is better than the current ontology of scientific change which doesn't include questions. To me, the answer to this latter question is a clear-cut "yes".
Generally speaking, a lack of a strict formal proof for the existence of a certain empirical phenomenon is not necessarily a serious reason for not accepting the existence of that phenomenon. As you know, I welcome the introduction of formalisms to scientonomy, but I urge all of us not to forget that we are not doing formal science here. Scientonomy is an empirical science and the acceptance of the existence of a certain empirical phenomenon doesn't necessarily require formal proofs. For instance, a biologist can accept that there are cows and dogs out there long before there are any formal proofs of it. A sociologist may legitimately speak of social institutions without any formal arguments for their existence. Similarly, the existence of questions can in principle be accepted without a proper formal proof. After all, we have never requested any formal proofs when accepting, say, Sebastien's ontology with normative theories. In short, I don't think that one needs to "formally treat the syntax of questions" before introducing them into the ontology of epistemic elements. Therefore, requiring "a more formal proof with explicit syntactic arguments" is something that is suitable to the fields of formal science, but is overly harsh in scientonomy. I don't think we have to stop doing empirical science until "all possible cases are considered". I believe this requirement is too stringent and would stale our science if it were to be employed. It would be reminiscent to the traditional approach taken by many philosophers, when the bar is set so unrealistically high that nothing ever gets accepted. This results in a stalemate where a criticism is followed by another wave of criticism and is trumped by more criticism. I think we would never move forward had we actually employed such harsh criteria. So I strongly urge scientonomists to stay away from this vicious practice and be more constructive – if a suggested modification improves our knowledge - albeit slightly - then it is to be accepted.
Now, is Patrick right to suggest that it is conceivable that questions can “be syntactically represented as a function of theories and methods”, i.e. that “question” can be somehow defined by means of "theory" and "method"? Yes, of course it is conceivable, simply because many things are conceivable. Yet, the fact is that we currently have no idea how this could be done. Will is making an excellent case that the extant attempts to reduce questions to theories are extremely dubious. I fully agree with him on that. I also don't see how questions can be reduced to methods. Finally, I don't think we have a slightest idea as to how questions can be expressed through theories and methods. Yes, Will doesn't cover this latter case, and Patrick is correct that this latter case should be studied in detail moving forward. But until we have any positive example of how questions can be properly expressed through theories and methods, the objection, albeit important, is not fatal for the suggested modification under scrutiny.
That being said, I encourage research into the formal aspects of Rawleigh's argument. I fully agree with Patrick that the formal proof is lacking. More research is needed to formally prove that questions are not reducible to theories and/or methods; I hope that scientonomists with a training in formal sciences will take on this important task going forward.
Yet, I strongly believe that a lack of a strict formal proof for the existence of a certain empirical phenomenon is not necessarily a serious reason for not accepting the existence of that phenomenon. Thus, my verdict is to accept Rawleigh's suggestion and introduce questions as part of our ontology, so that we start bridging the gap between our theoretical ontology and that of the encyclopedia and moving closer to implementing the database of intellectual history.
I agree with Hakob that formal, logical treatment, though desirable, is not required in empirical science, including in Scientonomy. He is also correct in pointing out that the sort of criticism that characterizes most philosophical discourse, if applied in empirical science, would thwart progress. As a result, even though it would be fantastic if Will manages to provide proofs that convince Patrick, they are altogether not needed for the acceptance of this modification.
I agree with Patrick that Will has not provided a convincing case that questions are irreducible to methods and theories. Unlike Patrick, who focuses mostly on the logical structure of Will's arguments, I make two criticisms that are more philosophical and empirical in nature. They point towards what I think are some major problems in his arguments.
Note: I have had some trouble in understanding how to add comments in the discussion tab. As I indicated, I am making two criticisms. I have added the first criticism here, and I will add the second one shortly after. Both of them should be read in order to properly understand my position, especially the 'Overall Thoughts' section after the second criticism.
Criticism 1: Irreducibility of questions to methods:
Rawleigh mentions that questions are irreducible to methods, because methods have an "inherently normative character" that "have the form of a hypothetical imperative" (Rawleigh, 2018, p. 4). Since a question cannot by definition be a hypothetical imperative, Rawleigh concludes that questions must be irreducible to methods merely due to logical syntax.
I find the first premise---that methods have an inherently normative character---problematic, because that is not the understanding/ definition of 'methods' accepted in scientonomy. For example, the currently accepted definition of method is "a set of requirements for employment in theory assessment" (http://www.s...i/com/Method). The definition of method in the context of a scientific mosaic is identical (http://www.s...ghyan-2015)). After the publication of Patton, Overgaard, Barseghyan (2017), the currently accepted definition of an employed method became "a method is said to be employed if its actual requirements constitute the actual expectations of the community" (http://scien...ghyan-2017)). Furthermore, the original formulation of a method read "a set of implicit rules to be employed in theory assessment", whereas methodology was defined as "a set of explicitly formulated rules for theory assessment" (Barseghyan, 2015, p. 53). The current definition of methodology is "a normative theory that prescribes the rules which ought to be employed in theory assessment" (http://www.s.../Methodology).
Given this evidence, I would like to make few observations. First, it is clear to me that methods, unlike the interpretation that Rawleigh advanced in his paper, do not have an "inherently normative character." On the contrary, they are merely sets of requirements that are employed in theory assessment. There is no reference to either the descriptive or the normative content of methods in the definition, let alone an assertion about their inherent character. (This is also the case with the definition of employed method.)
Second, notwithstanding the previous formulations of method and methodology, that were demarcated based mostly on their implicit and explicit requirements respectively, the accepted definition of methodology, unlike the definitions of method and employed method, clearly makes normativity its essential feature. This means that Rawleigh's argument about the irreducibility of questions to methodology may be acceptable. However, since Rawleigh claims that METHODS have an indispensable normative character, which precludes questions from being reduced to them, I cannot accept this argument as it fails to use the accepted definitions of method (and employed method). While I do not want to speculate, it may be the case that Rawleigh has unwittingly conflated method with methodology, and has used the definition of the latter in order to show how questions are irreducible to the former.
Criticism 2: Irreducibility of questions to theories
Rawleigh asserts that "the major step in interpreting the semantics [of the question] involved is shifting the content from the elements of the proposition to the status of the epistemic agent" (Rawleigh, 2018, p. 5). This leads to each question becoming a "description of the respective agent's mosaic at a specific moment of time" (p. 6), which is problematic because the semantic content of a question is dissimilar to the historical record of its acceptances by different agents at different times. The assumption enabling this argument is that "the semantic content of the question stays the same regardless of its historical fortunes" (p. 6). I think by 'historical fortunes' Rawleigh means the (positive) historical record of a question's acceptance.
I am unconvinced that the semantic content of the question stays the same irrespective of its context. I will give three reasons for doubting this claim. First, it is uncontroversial that many theories are proposed to answer a particular question, but to say that the question MEANS (i.e., semantic content) the same thing throughout different time periods is untenable. Take "why does the apply fall towards the Earth?" as an example. This question has been asked throughout history, and many attempts have been made to answer it. However, it had different meanings in different time periods. Whereas an Aristotelian would understand the "Earth" as the center of the universe, a Newtonian would characterize it as (I think) an oblate-spheroid. Likewise, in the question "what is human mind?", a Cartesian would say that it is an immaterial substance, while a contemporary psychologist or biologist would reduce it to neuro-physiological/ materialist conceptions of the brain.
This suggests that although agents may use the same terms to ask a question, their conceptions about the meaning of a question would be fundamentally different. To use classical terminology, different communities across times employ different fundamental taxonomy/ hard core concepts to ask questions. It is not at all clear that a question, though similar at the level of mere translation, would MEAN the same thing in different time periods. In other words, the referent is the same but the sense is different; unless I am mistaken, the semantic content is meaningfully understood as the sense of the term as opposed to its referent.
Second, although I agree with Rawleigh that questions are not reducible to propositions of theories due to the infinite regress argument, and I do not want to suggest a mechanism through which theories influence questions and vice versa, I think most scientonomists would agree that we intuitively think that theories and questions are somewhat related. At the very least, acceptance and rejection of theories make certain questions acceptable and others unacceptable; Rawleigh uses the example of the rejection of phlogiston theory to claim that certain questions became illegitimate forms of inquiry. He also suggests that questions and theories are connected, since they share many common features such as indicators of acceptance. Because we all agree that theories change through time, and if it turns out that questions are, perhaps only at a surface level, conjoined with and influenced by theories, it would be the case that when theories change, some parts of the (conjoined) question would necessarily change as well. Consequently, the semantic content of questions would not be ahistorical and would not stay "the same regardless of its historical fortunes."
I have a third criticism against this claim. I suspect most scientonomists would be familiar with theory-laddenness: The notion, understood at a rudimentary level, that theories influence the manner in which we interpret the world. This thesis supports my foregoing claims that "Earth" and "mind" mean different things in different time periods. This, in turn, leads me to think that one cannot meaningfully ask a question without interpreting the world with some theories. After all, what is it that one is asking the question ABOUT? Accordingly, by suggesting that the semantic content of questions is ahistorical, Rawleigh has altogether avoided the criticisms of the Quine-Duhem thesis, which formed an integral part of our scientonomic training. Therefore, we either need an argument suggesting that questions are somehow outside the purview of this thesis (which seems unlikely), or show that the Quine-Duhem thesis is mistaken (which seems difficult). It seems more likely to me that the semantic content of the questions is contingent upon the theories used to interpret the world, and ask questions about it.
Overall Thoughts:
Having made these criticisms, I must frame my comments in a more meaningful context. Although it is true that I find Will's attempts unsatisfactory, I am not arguing for the rejection of his proposed modification: I am neither arguing that questions are not distinct epistemic elements nor am I asserting that they are not irreducible to theories and methods. Rather, I am pointing towards some aspects of Will's arguments that I find untenable, and therefore, unacceptable. My claims are not attempted to repudiate his arguments in general. Indeed, it may be the case that I have completely misunderstood his work, and thus, my aforementioned criticisms are misguided. For that, I apologize and request that I should be pointed towards the right sections of the paper. I would highly appreciate comments from Will as well as other scientonomists about some or all of the claims I made here.
Will's other reasons---such as the infinite regress argument, the discrepancy between the semantic structure of the encyclopedia and our ontology of epistemic elements, and the status of questions in actual scientific practice---prove sufficient in convincing me that we should ACCEPT this modification. I am particularly convinced by his observation that there is a difference between the semantic structure of the encyclopedia and our scientonomic ontology, and the empirical argument that scientists do, in fact, ask questions that are, as epistemic elements, different from methods and theories. Perhaps, it may have been helpful to devote some parts of the paper towards providing historical examples that illuminated the distinction between questions, and theories and methods. Overall, I find his arguments very reasonable, and think that they open areas of further research, which will expand scientonomic knowledge.
A quick update on the state of this modification. We've had numerous discussions in the community concerning this modification and we have reached a consensus.
1. Patrick's point is taken: technically Will has shown that questions cannot be reduced to methods or to theories, but not to theories and methods. The question of whether questions can be expressed through some sort of a combination of theories and methods has not been discussed in the paper.
2. Because we don't currently have any idea as to how questions could potentially be reduced to/expressed through a conjunction of theories and methods, we currently accept questions as a separate epistemic element.
3. We are leaving the question on the Status of Questions open and encourage scientonomists to continue pursuing it. Specifically: are questions reducible to theories and methods?
By the community's decision, this concludes the discussion of this modification.
Enable comment auto-refresher
Patrick Fraser
Hakob Barseghyan
Ameer Sarwar
Ameer Sarwar
Hakob Barseghyan