1

What is the correct way to apply the prefix "non-" to negate a (maybe dashed) compound adjective?

Suppose that we want to negate a generic compound adjective "adjective1 adjective2". In this case:

  1. "non-adjective1 adjective2" looks a bit ambiguous since the scope of the prefix "non-" is at least unclear (in fact seems to affect only adjective1).
  2. "non-adjective1-adjective2" also looks unclear to me since the dashes seem to act at the same level, which is not true (yes, you can claim a precedence order from right to left, but I am not sure that this is canonical).
  3. So, the less ambiguous option should be something like "non-(adjective1 adjective2)" or "non-(adjective1-adjective2)", which may seem a bit too verbose.

Which of the previous options is more correct/convenient? Is there any general consensus (or better alternative) on how to negate compound adjectives?

Edit.- My question is not on whether or not to use double hyphens, the question is on how to make clear the scope of the negation of the prefix "non-". Normally my doubts arise in a mathematical context. An example could be "finitely generated" which, by the way, is an absolutely standard adjective in mathematics.

Edit2.- I just realized about a missing possibility which may be the best option:

  1. "non adjective1-adjective2"

In this case the scope of "non" is clearly the next word (which is what I want to negate) and is neat enough. Is this form (using the particle "non" as a separate word without hyphen) admissible?

Just to summarize for the given example "finitely generated" the options are:

  1. non-finitely generated
  2. non-finitely-generated
  3. non-(finitely generated) or non-(finitely-generated)
  4. non finitely-generated

I prefer the last one but I am not sure whether it is formally admissible.

suitangi
  • 139
  • 1
    It might help to provide an example of what you want to negate. Taking the example of "free range" as in chickens, I'm seeing a lot of "non-free range" but it may vary depending on the length of the expression, the degree of ambiguity, the familiarity of the phrase you're negating, etc. With something like "free range" everybody knows what it means and can infer that "non-free range" is the negation, but with another phrase it may be more ambiguous or confusing. – Stuart F Nov 22 '19 at 11:44
  • 1
    ... Yes, Grice rules here. Avoid (1b) ambiguity / (1a) any lack of clarity (I just had to re-order that, before the labelling disambiguated!) at all costs, and (2) outlandishness. Check in a dictionary first, to see if there is an established formatting. If you think you can't avoid these 'rules', rephrase. As @Stuart F wisely says, ' it [will] vary depending on the length of the expression, the degree of ambiguity, the familiarity of the phrase you're negating, [and potential outlandishness ... and hence target register]'. – Edwin Ashworth Nov 22 '19 at 12:06
  • The question is of a general nature (can I derive from your comments that there is no general rule?). Most of my examples come from technical (mathematical) context. An example could be "finitely generated" but there are many others (not necessarily with the same behavior with respect to negation). – suitangi Nov 22 '19 at 12:12
  • may I know why the downvote? – suitangi Nov 22 '19 at 12:31
  • just in case... @nigel I am not an expert in the English language and I don't know the exact meaning of "compound adjective", but "finitely generated" is certainly an adjective, and it is made of two words... – suitangi Nov 22 '19 at 12:36
  • 2
    I think someone here is non compos mentis. – Hot Licks Nov 22 '19 at 13:07
  • Both the accepted answer and the Chicago Manual of Style answer (https://english.stackexchange.com/a/78060/3306) are quite thorough. If that doesn't answer your question, perhaps you could ask another one? By the way, your question reminded of Gödel, Escher, Bach by Douglas Hofstadter. He carefully defined "self describing" and "non-self describing" in order to describe Gödel's Incompleteness Theorem. – rajah9 Nov 22 '19 at 14:02
  • I think the main problem with non finitely-generated is that it is viewed as wrong by many authorities. For example Merriam-Webster dictionary classifies non- as a prefix and not a word. So, even though lots of native English speakers use similar constructions, if you submit it to a journal, the copyeditors may try to fix your spelling. – Peter Shor Nov 23 '19 at 14:03
  • Thanks @PeterShor, I will take this into account. However, I am a bit surprised that there is no standard terminology for such a common concept. – suitangi Nov 23 '19 at 14:24

1 Answers1

0

If you do put a non- in front of a compound adjective, you should use two hyphens (or more, if needed): in your example, it should be non-finitely-generated groups.

But should you put non- in front of a compound adjective in the first place? This really depends on the example. Adding non- in front of a compound adjective can make it ambiguous; I would recommend only doing it if it's clearly non-ambiguous (like the first examples below).

There are some compound adjectives that sound perfectly fine if you add non- in front of them:

non-English-speaking customers,
non-nuclear-powered submarines.

There are some compound adjectives where the non- is ambiguous as to what it means:

non-Christian-oriented organziations.

Are these organizations oriented towards non-Christians, or are they organizations that are not specifically oriented towards Christians. You certainly shouldn't use ambiguous terminology.

Finally, there are compound adjectives where at first sight, the non- seems ambiguous, but some thought shows that there's only one reasonable interpretation:

non-blood-sucking insects.

A reader's first reaction might be, "What do they suck if it's not blood?" I wouldn't use these, either.

For the example you give, non-finitely-generated groups, I think it sounds fine. But there are probably lots of compound adjectives in mathematics you don't want to put a non- in front of; treat each case separately.

Peter Shor
  • 88,407
  • Thanks @Peter, my main concern about "non-finitely-generated" is that it does not make explicit the scope of "not-". Is it a general rule in English that the scopes of hyphens nest from right to left? – suitangi Nov 22 '19 at 13:16
  • Would it be OK to write "non finitely-generated" (without the first hyphen)? (In this case the scope of "non" is not ambiguous and precisely the desired one) – suitangi Nov 22 '19 at 13:23
  • If you want, you can use an en-dash after non, so non–finitely-generated, which specifies the order of nesting. This is the way the authorities say to do it (although in fact, many readers don't pay attention to the length of these hyphens). I wouldn't use non by itself. – Peter Shor Nov 22 '19 at 14:04