I saw a post on Programmers.StackExchange in which someone asks the question:

Why do programmers [or anyone] give the "dead fish" hand shake in an interview?

The responses to that question say things of the nature "not ALL programmers do that." The OP then replies, "I didn't say ALL programmers did that, I just said that programmers did that".

However, to me, it seems like saying "why do programmers do __" is equivalent to asking why all programmers do _____.

Who is correct? How many programmers give the "dead fish" hand shake, according to the phrasing of that question?


The absence of a qualifier like "most", "many", or "some" leaves the statement open to (mis)interpretation. Strictly speaking you are correct (and that would be my first reaction too); saying "programmers do _" is the same, syntactically, as saying that "computers emit heat" or "dogs chase cats" or "toddlers throw tantrums" -- in all cases there are exceptions, but the statement describes the normative case.

It's a bad idea to generalize about behaviors, so the original speaker should have added a qualifier.


It's a generalization of programmers, and generalizations are never completely accurate if taken literally.

  • 5
    All generalizations are false. – Robusto Aug 15 '11 at 21:43
  • "All generalizations are false" is also a generalization. Paradox. :) – Malcolm Aug 16 '11 at 17:37
  • Pssh. Does nobody read the source of other people's answers for fun anymore? – mmyers Aug 16 '11 at 18:07

I (and I would imagine most native English speakers) interpret a plural without a qualifier, by default, as having the implied qualifier 'all'. I see news organizations like the BBC do this all the time. 'Doctors say that fruit cures rabies' (when a group of 10 doctors had the finding from a study), 'Police want new powers to ban baseball bats' (when a small group of hardline police officers do), etc. It's an unfortunately poor style of writing that I think has penetrated increasingly into modern usage, often for the reason of brevity.

So it is technically ambiguous, but I would always lean towards the implied qualifier, 'all'. 'All programmers...'


An obvious case in which "all" is not implied concerns 'people.' As in,

Why do people give dead fish handshakes in job interviews?

In most contexts, one would probably assume that the speaker does not mean by the above that all people give dead fish handshakes, because that would be obviously false. So although the sentence is ambiguous, strictly speaking, there's still a clear best interpretation, as follows:

Why does anyone give dead fish handshakes in job interviews?

In other cases, however, the ambiguity causes interpretive problems. When 'people' above is replaced with 'programmers,' the possible semantic range of the sentence is greatly reduced, making the implied 'all' more plausible.

Nonetheless, we can be generous interpreters. Thinking in terms of predicate logic, the statement "programmers give dead fish handshakes" could mean either "there exist two or more programmers who give dead fish handshakes" or "all programmers give dead fish handshakes." These statements can be negated: "all programmers avoid dead fish handshakes" negates the former, while "some programmers avoid dead fish handshakes" negates the latter. So, interpreting generously, it seems likely that the OP meant something like

Why do any programmers give dead fish handshakes?

Or, I think more clearly in this case, its double negation:

Why don't all programers avoid dead fish handshakes?

This still raises the question of why the OP is talking about 'programmers' specifically, when the question applies equally well to all job interviewees. Perhaps the OP thinks that programmers, at least, should have the good sense to avoid dead fish handshakes -- still a generalization, but a more positive one, for programmers at least. Or perhaps the OP was just trying to ensure that the question was perceived as on-topic :)


All it really has to mean is that limp handshakes are significantly more common from programmers than people chosen from the general population (in the opinion of the person asking why they do it).

If, say, 1 in 4 programmers do this, but only 1 in 20 in the general population, that's more than enough to raise a legitimate question. I have no opinion on the actual numbers, of course.

It wouldn't be ridiculous, for example, to ask "Why do sharks attack people?", despite the fact that the percentage of all sharks that actually do this is vanishingly small. It's all relative.

  • "It wouldn't be ridiculous, for example, to ask "Why do sharks attack people?"" Well, this is because it only suggests that sharks may have a different reason for attacking people than other things that attack people, for example. "Some people are bad handshakers and some of those people are programmers" is singularly uninteresting, and the question implies an expectation there's more than that. – Random832 Aug 16 '11 at 6:03
  • @Random832: I don't understand what you mean. Assume any of your own (limited) experience, published survey results, or common belief suggest some subset of the general population are more inclined to do something than the average. Either that proposition is untrue (you, the survey, or people at large are mistaken), or there is some potentially interesting reason. It's irrelevant exactly how prevalent the unusual behaviour is - all that matters is it's noticeably more prevalent in the subset of interest. – FumbleFingers Aug 16 '11 at 17:01
  • If 0.001% of all people called "John Smith" spontaneously combust in any given year, but only 0.000000001% of people called anything else do this, I would start looking for an explanation. Perhaps some genetic condition associated with the Smith family - I don't know. Either there would be something significant about that subset, or it would be a very unlikely coincidence. – FumbleFingers Aug 16 '11 at 17:07
  • That's the point. It does imply there is a potentially interesting reason. If it's not actually more prevalent [nor any reason to suspect that there is a different reason than other populations, as for sharks] then there's no potentially interesting reason. It therefore doesn't have to imply "all" for the objections to the question to be valid. – Random832 Aug 16 '11 at 17:58
  • @Random832: I'm still not with you. We have to assume the person is asking the question because he has reason to think something is more prevalent among the subset he asks about. This has nothing to do with whether or not he's right in thinking that. Any response criticising the supposed implied "all" smacks to me of somewhat desperate defensiveness. – FumbleFingers Aug 16 '11 at 18:20

I think there may be a regional or cultural element to this as well. I grew up in New England, and I could say something like "New Yorkers are rude", and all the people I grew up with would understand that I am not speaking in absolute terms; the word "all" would NOT be implied in a sentence like that. It merely means "The majority of New Yorkers that I've met are rude". Yet in other parts of America, some folks think that the word "all" IS implied, as I've discovered. It amazes me to think that some folks think that I am saying that "I've personally met all 8 million New Yorkers and have found that every last one of them, without exception, is rude."!

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.