On Philosophy

June 19, 2007

Method: How To Avoid Writing Nonsense

Filed under: Metaphilosophy — Peter @ 12:00 am

As a philosopher the thing to fear is not being wrong. Being wrong is inevitable. The thing to fear is writing nonsense while being unaware that you are writing nonsense. This is not because nonsense doesn’t pay or isn’t respected. It is perfectly possible to be paid to write nonsense, and even to write nonsense well enough that other people begin to think of it as profound. So I guess if you do philosophy just as a job or to be some kind of public figure then writing nonsense might be fine. But philosophy generally doesn’t pay that well, and it is rare for a philosopher to become well known, so I suspect that if you do philosophy it is because you want to contribute to philosophy, to make some small progress of your own. In that case nonsense, even nonsense that convinces other people that it is not nonsense, is to be avoided at all costs. Obviously it is best to be right, but even being wrong can be a contribution to philosophy as a whole, by eliminating one solution from the list of possibilities. But nonsense makes no lasting contribution at all. Once it is recognized as nonsense it will simply be discarded.

The problem with nonsense is that there is no guarantee that it will be immediately recognized as such. Certainly it never seems to be nonsense to the author. And so the author will thus present it as making sense, and may do so well enough to convince others that it has content. And once the ball gets rolling it becomes harder and harder for people to come out and say that it is nonsense, especially when it starts being taught in classes. Instead of rejecting it as nonsense people begin to think that their lack of understanding is a sign that it is really deep and profound. This may lead them to study it until it begins to make sense to them, and which point they will begin to champion it as deep and profound, repeating the cycle. So, to get back to my original topic, even if we are writing nonsense there may be no clear indication of that fact. And so to avoid accidentally writing nonsense it is necessary to look closely and objectively at our own work.

But, before I describe a relatively reliable process for detecting nonsense, let me first describe the traditional rules of thumb for detecting nonsense.
1- most sentences contain at least one word that cannot be found in a small dictionary
2- it is hard to explain what is being claimed in simpler terms
3- an obsession with oneness, being, transcendence, or nothingness
4- capitalization of Words other than proper names in the middle of sentences
5- putting single “words” in quotes without sufficient reason
6- mathematical terms used in a non-mathematical context
7- others say that it sounds eastern or post-modern
Of course these rules of thumb aren’t perfect. If is perfectly possible for something that makes sense to violate some or all of these rules, and it is perfectly possible for nonsense not to break any of them. So these rules of thumb aren’t the best when it comes to evaluating our own work. However they can be quite useful when there is only a limited time in which to read a large number of journal articles, instead of carefully considering whether each particular paper makes sense we can save ourselves from reading many of them by applying these rules of thumb, at the cost of skipping the occasional useful paper disguised as nonsense.

Given that the traditional rules of thumb are imperfect I think that the best way to detect nonsense is simply by considering what the world would be like if the claim being considered was false, and some competing claim was true. If the world would be exactly the same, or the possibility of the claim being false is incoherent, then there is a good chance that the claim in question is nonsense. For example, consider the mystical claim that “all is one”. Would the world be different if the opposite, that “all is not one”, were the case? I can’t see how it would be. Thus it appears that the claim that “all is one” is not really claiming anything at all. It is thus nonsense, a non-claim verbally dressed up to seem like a claim. Or consider the claim that “all successful contemporary societies are virtuous”, where virtuous is defined as the property of not being conquered by an external foe. Considering what the world would be like if the opposite is true is to consider what a successful contemporary society that is conquered by an external foe would be like. But obviously that is a contradiction in terms, because it is impossible to be both a successful society and a conquered one. Thus the claim is revealed as not actually a proper claim. In contrast consider that same claim where virtuous is defined as having a mechanism for resolving internal disputes. Considering the opposite being true in this case is considering a world containing a successful contemporary society with no method for resolving internal disputes. Although such a society seems unlikely to exist it isn’t impossible by definition, and thus the claim is a real claim.

But despite the usefulness of this method there is one important exception to keep in mind, namely when dealing with claims that are intended to be definitions. Often a work that is not nonsense will want to provide some terms with more precise meanings, and will thus begin by defining those terms. Obviously such definitional claims are tautologies, and hence it makes no sense to consider what things would like if they were false. But, on the other hand, this is because they aren’t really claims at all, properly speaking, rather they provide the framework with which claims will be later made. So when evaluating whether something is nonsense it is important to remember to evaluate the central claims with this method, and not the preliminary definitions. (Of course something that is all definitions is nonsense, unless it is a dictionary, because it is probably portraying those definitions as if they were claims.)



  1. Not to trivialize your point (which I do agree with), but I can’t help but be amused at the restrained, carefully-thought-out way in which you’re almost lampooning a whole host of bad philosophical approaches.

    Comment by Philip L. Welch — June 19, 2007 @ 12:44 am

  2. I find it hard to take the subject seriously myself. The fact that many people treat such nonsense philosophy as profound alternately amuses and depresses me.

    Comment by Peter — June 19, 2007 @ 12:48 am

  3. Executive summary: “Continental philosophy blows.”

    Meh. Depends on what you’re trying to do. “All being one” combined with some other facts may have some importance for our value system. Ie. “Since all is one, I should treat other as myself” or whatever.

    This entry is very sweeping, and the argument for it is purely instrumental. However, instrumental arguments are inappropriate when the question at hand is, “what should we value?”

    Comment by Carl — June 19, 2007 @ 1:41 am

  4. How can something that has no consequences whatsoever, which the universe would be exactly the same were it to be false, have any imact on our value system?

    Comment by Peter — June 19, 2007 @ 1:49 am

  5. And not all continental philosophy is nonsense. Properly restricted you can say some interesting things in phenomenology, but only about the nature of experience as experienced. Its just that not many phenomenologists properly restrict themselves.

    Comment by Peter — June 19, 2007 @ 1:50 am

  6. “But obviously that is a contradiction in terms, because it is impossible to be both a successful society and a conquered one.” – this is somewhat debatable, even if I can’t think of a historic example it could be argued that one can imagine and describe a conquered yet successful society.

    Comment by moromete — June 19, 2007 @ 3:46 am

  7. What us conquered?

    Comment by Carl — June 19, 2007 @ 3:53 am

  8. Depends how you define sucessful. But in any case, why worry about the totally made-up example used as illustration? It’s not like anyone seriously holds that opinion.

    Comment by Peter — June 19, 2007 @ 12:53 pm

  9. Peter, I have to say I’m not entirely sure I understand the conclusion you’re trying to establish. The word “nonsense” seems ambiguous to me. Are you saying that “nonsensical” claims are necessarily false, that they are without truth-value, or that they have no meaning / content at all? Or are there different types of nonsense, some of which are necessarily false, others of which are without truth value, and still others of which have no meaning / content at all?

    Comment by Jason Zarri — June 19, 2007 @ 5:10 pm

  10. I am bypassing making those distinctions alltogether to make the claim that that is identified as nonsense here is bad philosophy / useless to philosophy. The nature of the truth-valueof the claims (which in the case of tautologies might very well be necesessarily true) is irrelvant. That might mean that we could proceed to classigy nonsense into further types, but I have to admit that I don’t see the value of such an exercise, unless we were doing philosophy of language. But what I have defined as nonsense here and what is nonsense in the context of language as a whole are not completely identical.

    Comment by Peter — June 19, 2007 @ 5:16 pm

  11. “If the world would be exactly the same, or the possibility of the claim being false is incoherent, then there is a good chance that the claim in question is nonsense.”

    Doesn’t this make this post nonsense?

    Also, is the position you are defending here logical positivism? Sure sounds like it…

    Comment by Richard Brown — June 21, 2007 @ 5:06 am

  12. Richard I don’t think it’s really positivism since he’s not arguing such statements are meaningless more that they are useless for philosophy. So he’s definitely not going as far as the positivists do. However he does appear committed to taking any claim in isolation relative to observables (as he discussed in his comments to my post) That does seem very similar to positivist methadology even if he’s not making claims about truth or meaning.

    I should add that the issue of verification or falsification relative to single observables or to systems or subsystems is a big deal in my book. I simply don’t think it coherent to speak of them relative to a single observable since inferences always need to be made to generate a difference. That always entails reference to multiple claims and never a single claim. (I’ll not argue that here)

    Comment by Clark — June 21, 2007 @ 11:40 am

  13. This post could be falsified by demonstrating how nonsense may slip through this criteria or by demonstrating a better way to catch nonsense.

    Comment by Peter — June 21, 2007 @ 12:24 pm

  14. I think the best way is to either ask the hermeneutical question (is this interpretable) or to ask the verification question (can this be verified or falsified in some broad sense of those terms) That’s why I said that in general I agree with you but just quibble on the details.

    The hermeneutic issue is sort of a broadening of your point about definitions. It captures both circularity that is typical of definition problems but also the problem of not even having definitions or reference. Verificationalism of a sort (not the positivist sort) then captures the problem of people making empty claims do to it making no difference.

    Comment by Clark — June 21, 2007 @ 1:29 pm

  15. I don’t think “is this interepretable” can quite serve for two reasons:
    1- I have never seen an objective way of determining what is and isn’t interpretatable, probably because of …
    2- Everything is interpretable with a complicated enough set of rules (anything can be mapped to anything in a systematic way as long as we are willing to let the mappings be as complicated as we like). Which leads the the conclusions that a) nothing is nonsense and b) everything means (can be interpreted) everything. Which I think makes a mess of truth.

    I assume, more simply, that the english language (or whatever language we are working with) already provides a single interpretation for all the terms, and that the way to interpret the peice is to let each sentance have its usual english meaning, deviating from that meaning only when words are explicitly re-defined. Of course I didn’t say that here, it’s more of a background assumption on my part. Given that, I am led to think that the only remaining question to ask is: do those statements actually make a claim? Or do they just pretend to? And I think that philosophy is supposed to be about the world and not about abstraction alone I am led to say that a claim must be a claim about the world. Which leads to phrasing that seems verificationist or falsificationist, but is really just me trying to express the idea that the claims really have to say something about the world, to restrict what is possible in some way, otherwise they aren’t saying anything about the world, but rather leaving it completely as it is.

    Comment by Peter — June 21, 2007 @ 1:45 pm

  16. I think interpretation and truth are somewhat different. After all some statements are interpretable but false. Now they are related of course, although not necessarily in a straightforward way. More in the way of understanding someone asserting something to be true – although even there I think we have to be careful as sympathetic as I am to Davidson.

    However to your main point I’d hope we’d not say that being overly complicated is to write nonsense. That would indeed be unfortunate. Now we can indeed say someone is a poor writer and we ought praise good writing in philosophy. But heaven knows there’s no shortage of bad writing in philosophy.

    The issue of interpretation is of course to raise the question of interpretable in terms of what code. (i.e. what framework that provides intelligibility) My sense is that you see this as a negative. (Perhaps unsurprising given your prior comments) I see this as a positive. (And frankly unescapable anyway)

    To claim that the English language provides interpretation for all terms seems problematic depending upon how broadly we interpret “English” or any other language. After all I can readily point you to some engineering manuals who use many terms in no general dictionary. So either you claim renders English far too broad to be of service or else excludes as nonsense most technical discussions.

    Now the way out is to allow that technical discussions have an implicit reference to other language use. Perhaps you’d require that in those places with implicit reference one needs terms explicitly defined. However I don’t think that works either since one might dispute what terms mean with there being no obvious definitions anywhere. This isn’t an insignificant matter either since arguably a lot of philosophical literature on just such problems in major philosophers can be found. The terms clearly aren’t being use in the typical sense. But it isn’t obvious what their sense is. Thus enabling many burgeoning philosophers to solve the publish or perish problem. This isn’t just a matter for understanding philosophers writing before the 20th century. (After all you might indeed consider Kant, Plato, Aristotle and so forth as actually writing nonsense) Rather I see no shortage of writings like this on major extremely well respected figures. Consider the controversy over say some of Davidson’s rather terse and sometimes ambiguous writings.

    Surely we don’t want to confuse ambiguity or poor writing with nonsense.

    Comment by Clark — June 21, 2007 @ 2:37 pm

  17. To add, I think the “is interpretable” ends up depending upon some sense of charity. After all if we can read someone in a fashion that makes sense, shouldn’t we? Isn’t an important aspect of reading philosophers especially those we disagree with is to read them as charitably as possible?

    Does this always result in a correct reading? Of course not. But I think it a fair demand to take when raising the spectre of nonsense.

    Comment by Clark — June 21, 2007 @ 2:41 pm

  18. One final point to anticipate an objection. I don’t think one can say charity leads to “any interpretation” as you seem to suggest. Clearly interpretation has to be grounded in reason, language and a concern for context. So everything doesn’t go as you claim.

    If something can’t be charitably interpreted as making sense then it is fair to call it nonsense.

    If something interpreted charitably makes sense but the meaning entails no possible difference in the world then it is fair to call it nonsense.

    Comment by Clark — June 21, 2007 @ 2:48 pm

  19. The problem I have with allowing such loose interpretation is that there is no one interpretation, and so there is no way to settle the truth of anything. Thus I think that much charity is too much charity (the point of communication is to be understood as intended; if no one can understand you as you have intended, especially if there is more than one possible interpretation; this then is a failure to communicate). As to code and technical manuals, surely you are aware that there is always a method of decoding / defintions for technical terms provided in plain english. That’s what makes them understandable / meaningful.

    In other words, I would rather classify poor writing as nonsense then allow works to be interpreted arbitrarily, or even loosely (and who is to say how loose is too loose?). Otherwise every criticism can be responded to simply by inventing a new way to interpret the work, and that is absurd. And it would make philsophical progress hard to impossible, you could never say that so and so is wrong, only that a specific interpretation of them is wrong. And then even that claim can itself be interpreted in ways that it is right and ways in which it is wrong. And so on, ad infinitum.

    Argument by analogy: in other disciplines where things actually do get settled interpretations beyond the simplest interpretation (English + explicit redifinitions + defintions that are part of the background of the field) are simply not accepted.

    Comment by Peter — June 21, 2007 @ 2:52 pm

  20. Once again I don’t see the lack of one interpretation as a flaw. More a feature of reality. Our assertions are typically general interpretations that include a level of vagueness. So when we try and pin them down we always have this problem of multiple interpretations.

    Now in practice there’s usually not that much play.

    But as I said we certainly shouldn’t say that ambiguity or vagueness entails nonsense, which appears to be what your criticism would entail.

    As to your analogy, if physics is the master of disciplines where things do get settled then what you say is simply not the case. There are many things accepted as meaningful that are open to multiple interpretations. Quantum Mechanics being the obvious example. Now one can attempt to argue that in these cases such matters aren’t really physics. But while you may convince a few of that I don’t think all physicists would agree in the least.

    Now certainly there are disciplines where there is much more agreement like physics. But if that is ultimately your model for philosophy then once again one must ask, how is this not just repeating the positivist project?

    Comment by Clark — June 21, 2007 @ 3:21 pm

  21. To add, it seems to me that the actual positivists are in a better position here than your “methadological positivism” is. (For lack of a better short hand term) After all the positivists allowed for a framework or something fulfilling that function so as to make observables meaningful. However by excluding reference to a system or subsystem you exclude this rather useful approach. So it’s not clear to me how your criteria of nonsense doesn’t end up the problem of self-refutation.

    Either it is true by definition and makes no difference in the world or else it is an observable claim without possible observation since you can’t appeal to a system. So either way it is nonsense.

    As I said the positivists had good answers here (if a bit problematic to some) This is why the typical facile refutations of the positivists fail. They actually had much stronger positions than their boogey-man status suggests. Now one could move from the positivists to their transformed form in the naturalized epistemologists like Quine. (Arguably Quine less refuted positivism than modified it in a few significant ways) But of course there you’ll have problem since Quine certainly would reject how you are treating observables as simple.

    Comment by Clark — June 21, 2007 @ 3:32 pm

  22. There is too much looseness. Just see the “all is one” example. (And postmodern philosophy in general in my opinion) If you aren’t willing to pin a claim like that down then it is too loose. Slight amibuity is unproblematic because on all acceptable readings the theory is true or false. But if interpretation can flip a false theory to being true or vice versa then that is too looseness, and becomes a obstacle to philosophy. If you object to A they rely on interpretation B. And if you object to B they go back to A. Or C.

    But the physics content of the theory, the predictions, is not open to interpretation. Similiarly the philosophical content of theory of philosopy should not be open to interpretation. Otherwise, again, we can’t go anywhere.

    I would rather be called a positivist, falsely, then be unable to make progress in philosophy, and I see opening the door to multiple interpretations as subverting the ability to make progress. We make progress by making a claim with a single meaning and then evaluating that claim. Again, allowing more than slight interpretation prevents anything from ever being settled.

    (See we haven’t really gotten into what I think makes a claim good or bad really, specifically that it explains well or poorly, and that is only related tangentally to observation. Suggestion: stop worrying about how I think claims should be evaluated.)

    Comment by Peter — June 21, 2007 @ 3:33 pm

  23. But Peter you explicitly denied the ability to pin such talk down by denying talk to a system. So you say “all is one” has to be readily interpretable and not in term of larger (and more complex) arguments such as in Plotinus.

    The issue after all isn’t simply vagueness but also the issue of interpretation (or hermeneutics) which entails an appeal to context.

    Note that I’m not calling you a positivist precisely because you aren’t engaging with these issues whereas the positivists did. I should also note that I think the positivists all too often get a bum rap: so when I’ve spoken of them in the above it’s always been in a positive fashion. (Excuse the pun) I disagree with them: primarily over this issue of vagueness. But I rather dislike their boogey-man status in both analytic and continental philosophy.

    Anyway, I originally raised the spectre of positivism not as a subtle ad homen but to try and get you to perhaps explain where your difference from the positivists lie. As I said my guess is that you reject frameworks and treat the critique as more a methodological issue rather than an outright critique of meaningfulness. Unfortunately I think that approach has serious problems that aren’t resolved by the merits of seeking a simple and univocal view of philosophical content.

    Comment by Clark — June 21, 2007 @ 3:43 pm

  24. It is perfectly possible to pin down such talk by simply providing defintions for “all” and “one”. With the meaning of the words pinned down (and the meaning of the words in the defintion, as far as is necessary) the claim is thus pinned down.

    As for the difference between myself and the positivists: we are simply doing different things, tackling different issues. We only overlap as much as I think that if we are making a claim about the world we must be saying something about the world. And to say something about the world requires is to make an assertion, somewehre, that things are such and such when they could have been so and so.

    Comment by Peter — June 21, 2007 @ 3:52 pm

  25. But that’s the problem. You want simple definitions whereas many (myself included) don’t think everything can be so simply defined.

    It just seems quite odd to say that complexity entails nonsense. As I said, folks like Plotinus may well have nonsense in them. They may well be wrong. But I don’t think it’s for the reasons you suggest.

    Comment by Clark — June 21, 2007 @ 4:01 pm

  26. Its not complexity; I admit that the defitions of all and one could very well be really complex, and in need of subsidiary defintions to clarify them (where have I said that I am against complexity?). But ultimately it has to bottom out somewhere, otherwise meaning and communication can’t get started (premise, otherwise where else can meaning get started from besides: language that already has meaning + inventing new words by non-linguistic definition).

    Comment by Peter — June 21, 2007 @ 4:05 pm

RSS feed for comments on this post.

Blog at WordPress.com.

%d bloggers like this: