falsifiability tossup and what to do in situations like this

[Auroni]

So, one of the editors packets featured this philosophy tossup:

10. Imre Lakatos questioned this idea by asking if Adams and Leverrier had not discovered Neptune. Carl Hempel used the Raven Paradox to challenge its generalizability. This principle is often demonstrated by noting the existence of black swans. Its formulator rejected the prevailing positivist and inductive views, positing that this principle applies because of the non-existence of a completely unique (*) science-specific methodology. A scientific theory meets this criterion if there is a reliable critical test through which a definitive counterexample can be found. For 10 points, name this principle that states that any scientific theory must be able to be disproven through experimentation, explicated by Karl Popper in The Logic of Scientific Discovery.
ANSWER: the principle of falsifiability [or word forms; accept equivalents like "the quality of being able to be falsified”; prompt on any phrase that mentions something like "the ability for a scientific theory to be disproven" before the last clause]

On the second and third sentences, there were many negs across multiple sites with "induction." Given the wording of those clues, I ruled in favor of accepting a buzz on induction (despite the fact that there was a separate tossup on the problem of induction elsewhere in the set) in a protest that I heard from the Minnesota site; at the Berkeley site, I instructed Tanay to outright accept a buzz of "induction" and to read a replacement for the other induction question. Needless to say, future sites will hear a different question than this one. But I wanted to post this in a separate thread to ask informed people about the following:

(1) Suppose that you have an ambiguous clue where someone says an answer that, after some research, is acceptable as well as what's on the page. Should the team get the points even if their answer does not apply to every clue prior to the one being buzzed upon? In this case, it doesn't seem as if "induction" should be acceptable for leadin buzz. However, I ruled that the points should be given because I felt as if a team should not be penalized for not knowing clues that they did not buzz upon.

(2) For questions on topics such as this, what's the best way to write the question without inviting negs from related topics? This seems like a really dumb and obvious question to ask, but I can see many editors in the future using a seemingly thorough and trustworthy source that discusses the raven paradox and the black swan example entirely with reference to falsifiability; the editor will then not think that another answer might also be at play. Is this something that has to be solved with playtesting involving the correct people, or is there a way to nip this at the bud?

[Red Panda Cub]

I think this question is especially tricky because Popper proposed his whole falsifiability shtick in large part as an attempt to nullify the problem of induction. The two concepts, therefore, are almost inextricably linked. Lots of things dealing with one will have strong connections with and connotations of the other. I'm not sure how to generalise that to a greater principle for question writing beyond hoping that question writers would take a bit more time to place the answerline within the greater context. Obviously that imposes a higher demand on the time of the writer/editor, but I think in all cases it is likely to produce a better question. Because of this, I'm not really sure that falsifiability is an great thing to tossup at all. At least, I'm not sure how to go about writing a full length falsification tossup without any clues that point to induction, though doing it the other way around is perfectly possible - as demonstrated in another packet from this tournament.

[theMoMA]

I strongly believe that questions as deeply flawed as this should be thrown out and a replacement in the same category read. The prevailing attitude in non-NAQT quizbowl seems to be that buzzes on ambiguous clues should be accepted. In this case, it seems to me that "induction" is simply wrong for the first clue, and definitely right (and probably more right than "falsifiability") on the second and third clues. In light of that, I don't see how it makes sense to accept an answer that is wrong for at least one of the clues. It definitely wouldn't be fair to allow a possibly game-changing neg to stand, so a new question definitely has to be read, but I don't think accepting the answer is good practice.

[Cheynem]

I agree with Andrew as I think it's just the simplest solution.

[Cody]

(2) For questions on topics such as this, what's the best way to write the question without inviting negs from related topics? This seems like a really dumb and obvious question to ask, but I can see many editors in the future using a seemingly thorough and trustworthy source that discusses the raven paradox and the black swan example entirely with reference to falsifiability; the editor will then not think that another answer might also be at play. Is this something that has to be solved with playtesting involving the correct people, or is there a way to nip this at the bud?

I'm not sure I understand this; whenever you write a question you have to vet the clues to make sure they are unique. You can't assume they are unique based on your experience with one source or another. That's the solution to nipping this in the bud.

[theMoMA]

I'll also add that it seems wrong to me to reward the first person to buzz in and say a plausibly correct answer when the other team may have been reticent to buzz because of misleading or ambiguous clues. I think the rule should recognize that both teams are equally disadvantaged by a tossup that has clues referring to two different answers, and should throw out such questions as a result.

[Auroni]

Okay, I suppose that seems fair. Basically, I opened this thread because I didn't know what the orthodox method of handling emergency situations such as this was. I was under the general impression that you should only throw questions out as a last resort.

Cody: I understand, but does this mean that you should use multiple sources to write every and edit every tossup? That seems impractical, given the demands and constraints of time.

[Cody]

Cody: I understand, but does this mean that you should use multiple sources to write every and edit every tossup? That seems impractical, given the demands and constraints of time.

It depends. Usually a quick google search of the clues will verify whether they are unique (and they are clearly unique in a lot of cases, so you might not have to worry about that). It's usually not too time consuming: for example, one could find that the black swan argument applies to the problem of induction in like ten seconds flat. Overall, it's not very impractical and does help a lot.

[grapesmoker]

10. Imre Lakatos questioned this idea by asking if Adams and Leverrier had not discovered Neptune. Carl Hempel used the Raven Paradox to challenge its generalizability. This principle is often demonstrated by noting the existence of black swans. Its formulator rejected the prevailing positivist and inductive views, positing that this principle applies because of the non-existence of a completely unique (*) science-specific methodology. A scientific theory meets this criterion if there is a reliable critical test through which a definitive counterexample can be found. For 10 points, name this principle that states that any scientific theory must be able to be disproven through experimentation, explicated by Karl Popper in The Logic of Scientific Discovery.
ANSWER: the principle of falsifiability [or word forms; accept equivalents like "the quality of being able to be falsified”; prompt on any phrase that mentions something like "the ability for a scientific theory to be disproven" before the last clause]

The real problem of this question is that about half of it just blows. I'm trying and failing to understand what is even meant by the first sentence; I haven't succeeded in tracking down the exact cite, but the way this is phrased makes zero sense. Come on, just read it aloud to yourself: "Lakatos questioned this idea by asking if Adams and Leverrier had not discovered Neptune." That's just barely a sentence even. All I can reasonably tell from that is that it's something to do with the philosophy of science. The second clue is as misleading as people have implied, and furthermore seems to me to be incorrect. What Hempel is discussing in his formulation of the Raven Paradox is something he calls Nicod's criterion of confirmation; he also explicitly distinguishes confirmatory and disconfirmatory evidence from falsification and verification. The third sentence I think is just filler. I suppose one could demonstrate falsification via the existence of black swans, but falsification is not "often" demonstrated this way (often by whose criteria?) nor does that sentence really convey a great deal of philosophical information. I guess one could plausibly buzz there with "falsification" but I wouldn't necessarily chance it. After this, the question goes on to the standard Popperian stuff which is fine, but everything prior to this is just either confusing or vague. If you wanted to work Lakatos into it, you could talk about the distinction he draws between naive and sophisticated forms of falsification in his "Falsification and the Methodology of Research Programmes;" that's something that would make sense and would be unambiguous. And the way to work Hempel into this question is not by just name-dropping the Raven Paradox without making sense of what it's actually about.

So basically, the problems with this question could be fixed by carefully reading the original sources and accurately rendering the arguments therein. This doesn't answer the question of "how to best resolve the problem of having written a bad question after the fact," but it does solve the "how to avoid writing that bad question in the first place" problem.

[PeterB]

10. Imre Lakatos questioned this idea by asking if Adams and Leverrier had not discovered Neptune. Carl Hempel used the Raven Paradox to challenge its generalizability. This principle is often demonstrated by noting the existence of black swans. Its formulator rejected the prevailing positivist and inductive views, positing that this principle applies because of the non-existence of a completely unique (*) science-specific methodology. A scientific theory meets this criterion if there is a reliable critical test through which a definitive counterexample can be found. For 10 points, name this principle that states that any scientific theory must be able to be disproven through experimentation, explicated by Karl Popper in The Logic of Scientific Discovery.
ANSWER: the principle of falsifiability [or word forms; accept equivalents like "the quality of being able to be falsified”; prompt on any phrase that mentions something like "the ability for a scientific theory to be disproven" before the last clause]

The real problem of this question is that about half of it just blows. I'm trying and failing to understand what is even meant by the first sentence; I haven't succeeded in tracking down the exact cite, but the way this is phrased makes zero sense. Come on, just read it aloud to yourself: "Lakatos questioned this idea by asking if Adams and Leverrier had not discovered Neptune." That's just barely a sentence even. All I can reasonably tell from that is that it's something to do with the philosophy of science. The second clue is as misleading as people have implied, and furthermore seems to me to be incorrect. What Hempel is discussing in his formulation of the Raven Paradox is something he calls Nicod's criterion of confirmation; he also explicitly distinguishes confirmatory and disconfirmatory evidence from falsification and verification. The third sentence I think is just filler. I suppose one could demonstrate falsification via the existence of black swans, but falsification is not "often" demonstrated this way (often by whose criteria?) nor does that sentence really convey a great deal of philosophical information. I guess one could plausibly buzz there with "falsification" but I wouldn't necessarily chance it. After this, the question goes on to the standard Popperian stuff which is fine, but everything prior to this is just either confusing or vague. If you wanted to work Lakatos into it, you could talk about the distinction he draws between naive and sophisticated forms of falsification in his "Falsification and the Methodology of Research Programmes;" that's something that would make sense and would be unambiguous. And the way to work Hempel into this question is not by just name-dropping the Raven Paradox without making sense of what it's actually about.

So basically, the problems with this question could be fixed by carefully reading the original sources and accurately rendering the arguments therein. This doesn't answer the question of "how to best resolve the problem of having written a bad question after the fact," but it does solve the "how to avoid writing that bad question in the first place" problem.

I think the first sentence makes quite a lot of sense. The example of the discovery of Neptune comes up often in the refutation of falsificationism, though more often with reference to the Duhem-Quine problem (which really ought to have been an early/middle clue, to make the toss-up clearly uniquely identifying) because according to Popperian falsificationism the observations in the orbit of Uranus which led to the postulation of Neptune ought instead to have been used to falsify Newtonian mechanics, which obviously would have been ridiculous. (Except then there was Mercury's orbit, which succeeded in doing that, but that's beside the point). It has relatively little to do with induction. I agree that naming Lakatos at this point is unhelpful except to put us in the area of Philosophy of Science. The two clues after that aren't great, but I don't feel they point to induction in any meaningful way. I mean, the black swan is clearly supposed to be referencing the idea that a black swan falsified the theory that all swans are white.

Anyway, I'm of the opinion that induction is simply wrong and should be counted as a neg, which is what happened in the final of the Oxford mirror which I was moderating. I was quite surprised at the time, given that induction had come up earlier for us, and I didn't really think the question pointed strongly in that direction. However, I concede that it is a badly-written question, given that so many people are negging on it.

[grapesmoker]

I think the first sentence makes quite a lot of sense. The example of the discovery of Neptune comes up often in the refutation of falsificationism, though more often with reference to the Duhem-Quine problem (which really ought to have been an early/middle clue, to make the toss-up clearly uniquely identifying) because according to Popperian falsificationism the observations in the orbit of Uranus which led to the postulation of Neptune ought instead to have been used to falsify Newtonian mechanics, which obviously would have been ridiculous. (Except then there was Mercury's orbit, which succeeded in doing that, but that's beside the point). It has relatively little to do with induction. I agree that naming Lakatos at this point is unhelpful except to put us in the area of Philosophy of Science. The two clues after that aren't great, but I don't feel they point to induction in any meaningful way. I mean, the black swan is clearly supposed to be referencing the idea that a black swan falsified the theory that all swans are white.

I think you're confusing the ability to connect the clues to what's being asked post hoc with whether the clues actually make sense. I mean, I understand perfectly well what role Neptune's discovery plays in the falsification debate; the problem is that this question makes a vague handwave in the direction of the problem and then goes "well, figure it out." That's not good question writing (to say nothing of the fact that the first clue is also a sentence fragment, making it not good writing, period).

[Cody]

The two clues after that aren't great, but I don't feel they point to induction in any meaningful way. I mean, the black swan is clearly supposed to be referencing the idea that a black swan falsified the theory that all swans are white.

Players are not supposed to read the question writer's mind. It doesn't matter what you think it is "clearly supposed to reference" when it is, in fact, not clear what it means at all. Any yokel can easily combine a meager understanding of philosophy and the English language to see that nothing the existence of black swans is often used to demonstrate the problem of induction and thus does point to induction in a meaningful way.

[PeterB]

The two clues after that aren't great, but I don't feel they point to induction in any meaningful way. I mean, the black swan is clearly supposed to be referencing the idea that a black swan falsified the theory that all swans are white.

Players are not supposed to read the question writer's mind. It doesn't matter what you think it is "clearly supposed to reference" when it is, in fact, not clear what it means at all. Any yokel can easily combine a meager understanding of philosophy and the English language to see that nothing the existence of black swans is often used to demonstrate the problem of induction and thus does point to induction in a meaningful way.

But equally "any yokel" with "a meagre understanding of philosophy" probably shouldn't be buzzing in on the second line of a philosophy question on a hunch. Also, the principle of induction isn't "demonstrated" by the existence of black swans (which is what the question said, not the "problem" of it); in fact, it's downright refuted! While the words "black swan" might point you in the direction of induction, that's a really bad argument against a question - listening to half the words in a clue and then guessing is not how one ought to be playing quizbowl, so saying that some people do that is not really relevant.

I agree that the early part of the question, particularly the second and third clues, is not good. (I didn't write the question, by the way, so I have no interest in defending it other than that it's in my field of study right now). My main point is that induction is still clearly an unacceptable answer for the question and that it shouldn't be accepted at all: the first line is a clear (and in my opinion, perfectly comprehensible) reference to falsificationism, the second clue is perhaps vague and not really uniquely identifying, but not really enough to buzz in on, and the third clue is actually pointing to the opposite of induction. While the selection of clues may be misleading and I'm certainly not going to use it as an example of a good question, I think that saying it's a case where induction should be accepted is definitely wrong. There were more plausible complaints about answer lines on other questions in the rooms I was moderating in (something about resonance springs to mind, but I don't remember the specifics now).

[Cheynem]

I'm not a philosopher, but it was explained to me that the buzz of "induction" implies "the problem of induction," so the black swan thing demonstrates "the problem of induction," not that "induction is true," which I agree is totally wrong. I don't think people who negged it were guessing but were trying to figure out what the question wanted; the ravens and black swan things are notable examples of philosophy, but I think it's difficult to tell whats going on. I can see the arguments for not accepting induction and just throwing the question out, but it's a problematic question.

[i never see pigeons in wheeling]

10. Imre Lakatos questioned this idea by asking if Adams and Leverrier had not discovered Neptune. Carl Hempel used the Raven Paradox to challenge its generalizability. This principle is often demonstrated by noting the existence of black swans. Its formulator rejected the prevailing positivist and inductive views, positing that this principle applies because of the non-existence of a completely unique (*) science-specific methodology. A scientific theory meets this criterion if there is a reliable critical test through which a definitive counterexample can be found. For 10 points, name this principle that states that any scientific theory must be able to be disproven through experimentation, explicated by Karl Popper in The Logic of Scientific Discovery.
ANSWER: the principle of falsifiability [or word forms; accept equivalents like "the quality of being able to be falsified”; prompt on any phrase that mentions something like "the ability for a scientific theory to be disproven" before the last clause]

The real problem of this question is that about half of it just blows. I'm trying and failing to understand what is even meant by the first sentence; I haven't succeeded in tracking down the exact cite, but the way this is phrased makes zero sense. Come on, just read it aloud to yourself: "Lakatos questioned this idea by asking if Adams and Leverrier had not discovered Neptune." That's just barely a sentence even. All I can reasonably tell from that is that it's something to do with the philosophy of science. The second clue is as misleading as people have implied, and furthermore seems to me to be incorrect. What Hempel is discussing in his formulation of the Raven Paradox is something he calls Nicod's criterion of confirmation; he also explicitly distinguishes confirmatory and disconfirmatory evidence from falsification and verification. The third sentence I think is just filler. I suppose one could demonstrate falsification via the existence of black swans, but falsification is not "often" demonstrated this way (often by whose criteria?) nor does that sentence really convey a great deal of philosophical information. I guess one could plausibly buzz there with "falsification" but I wouldn't necessarily chance it. After this, the question goes on to the standard Popperian stuff which is fine, but everything prior to this is just either confusing or vague. If you wanted to work Lakatos into it, you could talk about the distinction he draws between naive and sophisticated forms of falsification in his "Falsification and the Methodology of Research Programmes;" that's something that would make sense and would be unambiguous. And the way to work Hempel into this question is not by just name-dropping the Raven Paradox without making sense of what it's actually about.

So basically, the problems with this question could be fixed by carefully reading the original sources and accurately rendering the arguments therein. This doesn't answer the question of "how to best resolve the problem of having written a bad question after the fact," but it does solve the "how to avoid writing that bad question in the first place" problem.

To be clear, I wrote this question, and I definitely think that I made it far too vague in the interest of keeping it within 7 lines. I should've read this aloud to myself prior to putting it in the packet, but the sentence fragment at the beginning was intended to be "Imre Lakatos questioned this idea by presenting a hypothetical scenario where Adams and Leverrier had not discovered Neptune." And the example of the Raven Paradox is also flawed perhaps because I didn't make it clear that the paradox challenged the notion that the principle of falsifiability itself cannot be generalized: i.e. once you say "all ravens are black" from the process of inductive reasoning after observing thousands of black ravens, then the original statement is unfalsifiable in practical terms because it is impossible for an ornithologist to go around and study every single raven in the world to check whether it's black or not. But that was not part of the original Raven Paradox but rather an extrapolation made by other authors, so that's my fault for not making sure that the part pertaining to falsifiability was Hempel's work. The original phrasing of the second part was "a notable example used to demonstrate this principle is the historical premise that all swans are white" before Auroni changed it in editing (not to cast any blame on him for the poor question). Again, a very vague choice of words on my part and too much expectation of question-writer mind-reading. I supposed that people who knew about the swan example through a philosophy of science class would learn about it in a lesson about falsifiability, but they could just as easily learn about it as part of inductive reasoning.

[Cody]

But equally "any yokel" with "a meagre understanding of philosophy" probably shouldn't be buzzing in on the second line of a philosophy question on a hunch. Also, the principle of induction isn't "demonstrated" by the existence of black swans (which is what the question said, not the "problem" of it); in fact, it's downright refuted! While the words "black swan" might point you in the direction of induction, that's a really bad argument against a question - listening to half the words in a clue and then guessing is not how one ought to be playing quizbowl, so saying that some people do that is not really relevant.

What are you talking about? This has nothing to do with hunches or listening to only half the words in a clue! The question clearly states that 'noting the existence of black swans is often used to demonstrate this principle'. If you know what the black swan argument is (which I assume you do), you know this most directly points to the problem of induction. Given the inclusion of "often," you'd also be crazy to leap to falsifiability over the problem of induction. Since the question wanted an "idea" or "principle," the next logical step is to assume it wants "induction." There is nothing contradictory about saying that principle of induction is demonstrated by the black swan argument as this does not imply anything about the consistency (or lack thereof) of inductive reasoning.

I agree that the early part of the question, particularly the second and third clues, is not good. (I didn't write the question, by the way, so I have no interest in defending it other than that it's in my field of study right now). My main point is that induction is still clearly an unacceptable answer for the question and that it shouldn't be accepted at all: the first line is a clear (and in my opinion, perfectly comprehensible) reference to falsificationism, the second clue is perhaps vague and not really uniquely identifying, but not really enough to buzz in on, and the third clue is actually pointing to the opposite of induction. While the selection of clues may be misleading and I'm certainly not going to use it as an example of a bad question, I think that saying it's a case where induction should be accepted is definitely wrong. There were more plausible complaints about answer lines on other questions in the rooms I was moderating in (something about resonance springs to mind, but I don't remember the specifics now).

Please reread Jerry's last post about post-hoc analysis and his post before that.

[grapesmoker]

I don't think "induction" is the right answer either, based on some of the clues, but the whole point here is that the confusing wording might lead you to think that induction was being talked about.

I learned, having written a few not-so-great questions in this vein of my own, that to write questions on philosophical ideas well, you really need to focus on the meat of the idea. Trying to pack in a lot of clues is potentially harmful because you end up shortening useful information to try and cram more stuff in and not wind up with a 12-line tossup. My suggestion is that a good way forward is to track down a few solid sources, really try and understand what they're saying, and then try to some up their argument succinctly but fully. Obviously that's not a trivial task, but hey, that's why you write questions.

[PeterB]

But equally "any yokel" with "a meagre understanding of philosophy" probably shouldn't be buzzing in on the second line of a philosophy question on a hunch. Also, the principle of induction isn't "demonstrated" by the existence of black swans (which is what the question said, not the "problem" of it); in fact, it's downright refuted! While the words "black swan" might point you in the direction of induction, that's a really bad argument against a question - listening to half the words in a clue and then guessing is not how one ought to be playing quizbowl, so saying that some people do that is not really relevant.

What are you talking about? This has nothing to do with hunches or listening to only half the words in a clue! The question clearly states that 'noting the existence of black swans is often used to demonstrate this principle'. If you know what the black swan argument is (which I assume you do), you know this most directly points to the problem of induction. Given the inclusion of "often," you'd also be crazy to leap to falsifiability over the problem of induction. Since the question wanted an "idea" or "principle," the next logical step is to assume it wants "induction." There is nothing contradictory about saying that principle of induction is demonstrated by the black swan argument as this does not imply anything about the consistency (or lack thereof) of inductive reasoning.

I agree that the early part of the question, particularly the second and third clues, is not good. (I didn't write the question, by the way, so I have no interest in defending it other than that it's in my field of study right now). My main point is that induction is still clearly an unacceptable answer for the question and that it shouldn't be accepted at all: the first line is a clear (and in my opinion, perfectly comprehensible) reference to falsificationism, the second clue is perhaps vague and not really uniquely identifying, but not really enough to buzz in on, and the third clue is actually pointing to the opposite of induction. While the selection of clues may be misleading and I'm certainly not going to use it as an example of a bad question, I think that saying it's a case where induction should be accepted is definitely wrong. There were more plausible complaints about answer lines on other questions in the rooms I was moderating in (something about resonance springs to mind, but I don't remember the specifics now).

Please reread Jerry's last post about post-hoc analysis and his post before that.

I mistyped in my last paragraph and meant to say it's not an example of a good question, which I figure changes the meaning of what the post actually said. (I've now changed the original post) I'm not trying to defend the question as a good question, and I agree with a lot of the criticism and that the clues aren't meaningful.

What I disagree with is that induction could at any point be taken as a correct answer. You didn't really answer my point about swans: I don't really understand what principle you think is being demonstrated by the existence of black swans - I don't think you can realistically interpret "principle" as "problem". Unless the principle in question is "induction is bad", which would be a really bad answer-line, then you're still not really making much sense on that count, since black swans are never used to demonstrate that inductive reasoning is good.

On Jerry's point, I've already said I don't think the clues are good and I agree that some of the question is badly-worded and unclear. While reading, my thought was that I would have been sitting on falsificationism, albeit slightly confusedly, but wouldn't have buzzed until the bit about rejecting inductivist views. I don't think, however, that you can justify buzzing in on induction at any point.

[Cheynem]

The idea that "induction is bad" (i.e.. _problem of induction_) seems like an unorthodox but valid answer line. You're right that it doesn't make a lot of sense with the first clue, but in a game situation players shouldn't have to do extra amounts of mental leg work. Again, I can see the arguments for accepting it at a point and throwing the question out period, but it would be very problematic to just neg someone for saying induction.

[grapesmoker]

What I disagree with is that induction could at any point be taken as a correct answer. You didn't really answer my point about swans: I don't really understand what principle you think is being demonstrated by the existence of black swans - I don't think you can realistically interpret "principle" as "problem". Unless the principle in question is "induction is bad", which would be a really bad answer-line, then you're still not really making much sense on that count, since black swans are never used to demonstrate that inductive reasoning is good.

Ok, the problem here is that you're not looking at the original formulation of the clue; you're addressing some other, better clue that might have been used and employed black swans. Here's the original sentence: "This principle is often demonstrated by noting the existence of black swans." That's it. There's no sense here of how the existence of black swans demonstrates the principle of falsification and that's the key missing ingredient. It's the "this effect is associated with X" clue of philosophy, because even though the reasonably plausible thing to do here is to say "falsification," there's no real way you can be sure. "Induction" is a plausible answer here, simply because inductive reasoning could lead you to conclude falsely that all swans are white. Hell, "Bayesian updating" is potentially a right answer here too!

If there's a universal rule that needs to be taken away from this question, it's that you cannot just kind of wave in the direction of a clue and expect people to figure it out. This is wrong and bad and you shouldn't do it. If you can't explain precisely what you mean with a particular clue, then you shouldn't use that clue. That's why in science when people say things like "X is associated with Y" people who actually know science get upset; it renders the clue unintelligible.