Talk:free variable

The following information passed a request for deletion.

This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.

---

SoP.​—msh210℠ 19:36, 20 July 2009 (UTC)

Is it not a technical term in mathematics? Certainly the current definition is well beyond WT:CFI, so delete unless it can be attested as a specific technical term, not a variable which is free. Mglovesfun (talk) 20:08, 20 July 2009 (UTC)

It is just a variable which is free, q.v.​—msh210℠ 20:12, 20 July 2009 (UTC)

Keep.

With the current definition from programming, the term "free variable" appears to be SoP only because its meaning is explicitly listed in the entry "free" -- "(programming) Of identifiers, not bound". The same applies to "free variable" in logic, which is currently undefined.

If "free variable" gets deleted, other terms may follow. They include algebraic number, per the definition of algebraic -- "(Of a number) which is a root of some polynomial", which makes "algebraic number" technically a sum of parts. Likewise transcendental number and even complex number, as complex has the definition "(mathematics) Of a number, of the form a + bi, where a and b are real numbers and i is the square root of −1."

I fear that these cases provide a method of how to artificially make a lot of two-word technical terms of the form <adjective> <noun> appear sum-of-parts, by providing their definition at the adjective, of the form "Of <noun>, definition". Imagine I get rid of red dwarf by adding to red the definition "Of a dwarf star, small and relatively cool one of the main sequence".

I do not know what WT:CFI says to these cases, but to me all these sum-of-parts seem somehow artificial or odd. I would like to see free variable, algebraic number, transcendetal number and complex number included.

Some of the concerned entries: algebraic number, algebraic integer, bound variable, cardinal number, complex number, free variable, imaginary number, rational number, real number, transcendental number, free software, open set, closed set, complete graph, normal distribution. --Dan Polansky 22:51, 20 July 2009 (UTC)

See especially prime number, where this discussion already happened. Equinox ◑ 02:10, 21 July 2009 (UTC)

Yeah, I thought it was time to revisit the issue.  :-)  (The previous discussion is at talk:prime number.)​—msh210℠ 18:21, 21 July 2009 (UTC)

But note, Dan (and others), that people speak of a variable's being free, without tying the word free into the phrase free variable: google books:"variable is free".​—msh210℠ 18:21, 21 July 2009 (UTC)

Yes, the adjective and the noun is separable, not glued together, in most of the listed cases.

For informal comparison outside of bounds of WT:CFI, many general dictionaries have "prime number"[1] and "complex number"[2], while only few general dictionaries have "free variable"[3]. --Dan Polansky 22:45, 21 July 2009 (UTC)

I have no doubt that it's important to include prime number, and I agree with Dan Polansky's reasoning. But, for free variable, I don't know. Lmaltier 18:58, 21 July 2009 (UTC)

I am always looking to break phrases down to components, but I don't see the point in the case of well-defined terms like this and many other mathematical and scientific terms. The parallels among, say, logical, mathematical, computing, and linguistic senses seem real, but each use of "free" is quite distinct and doesn't occur except in close proximity and obvious reference to "variable". I would think we could make a CFI argument for this. Frankly, I'd even prefer not to try to do the forced one-collocation, one-context definitions at "free". Wouldn't it make more sense to have some sense at free that accentuated the parallels and directly referred users to the entry at free variable which contained the context-specifics? DCDuring TALK 00:49, 22 July 2009 (UTC)

The math/logic sense of free for variables is used in reference to variables only, of course, but not always with the word variable. E.g., google books:"is free in the statement|predicate|formula" -"variable is free" (some of which do use variable, but many, many of which do not).​—msh210℠ 17:18, 22 July 2009 (UTC)

Perhaps subsenses s.v. free-?​—msh210℠ 17:18, 22 July 2009 (UTC)

Weak keep, I did ask for this entry to be cleaned up into "common English", but since nobody's proposed bound variable as SoP I don't see why this should be deleted. Mglovesfun (talk) 11:53, 9 August 2009 (UTC)

Kept for no consensus.--Jusjih 03:44, 25 December 2009 (UTC)

The following information passed a request for deletion.

This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.

---

free variable

See #countable noun above. Someone mentioned this as precedent but I believe this is the same. If you can say "that variable is free" then "free variable" is no longer a fixed expression, but it is free that has a specific sense when referring to a variable. —CodeCat 14:06, 10 November 2012 (UTC)

"No consensus" isn't a very solid precedent. Still, I think it is better to keep this due to the myriad meanings of free, the applicable one not something that I'd likely use outside of this phrase as the meaning would be unclear. DAVilla 03:25, 11 November 2012 (UTC)

• Keep. I was mentioning "free variable" in another RFD, but not really as a precedent. If anything would be a predecent, it would be Talk:prime number. I was mentioning Talk:free variable as a page where I have argued the case for a whole class of terms. The class includes algebraic number, algebraic integer, bound variable, cardinal number, complex number, free variable, imaginary number, prime number, rational number, real number, transcendental number, free software, open set, closed set, complete graph, normal distribution. In these terms, the adjective is separable from the noun (as in "The number is prime"), but the used sense of the adjective is specific to the noun. The italicized clause of the previous sentence does not hold of "brown leaf" and "green hat", prototypical sum-of-part terms. In my mind, I store the meanings of these terms on these terms as listed above; an entry in my mind (as it were) is on "algebraic number" rather than on "algebraic"; it is on "free variable" rather than on "free". Thus, I naturally look up "algebraic number", and expect other users to do the same. Some dictionaries do have some of the terms listed above, but not all of the terms; see “prime number”, in OneLook Dictionary Search. and “complex number”, in OneLook Dictionary Search.. For dictionaries having "free variable", see http://mathworld.wolfram.com/FreeVariable.html, and http://dictionary.reference.com/browse/free+variable?r=66. As to whether the terms listed above are semantic sum of parts: they are if you equip the adjective with a definition specific to the noun. But this is a consequence of separability. A somewhat related case seems to be the one of "look up", which can be used in "look it up", yet having "look up" in a dictionary is essential. --Dan Polansky (talk) 10:53, 11 November 2012 (UTC)

• Comment. Contrary to what Dan Polansky suggests, this sense of "free" is not specific to the noun "variable". Consider, e.g., “Notice that ‘y’ is bound in the formula, and that only ‘x’ is free” (here). google books:"free occurrence" and "free occurrences" both get thousands of hits. And even with a query that does allow it to modify “variable”, namely "occurs free", I find that only in a small minority of cases does it genuinely do so. Also contrary to what he suggests, MathWorld is not exactly a "dictionary". —Ruakh_TALK 11:17, 11 November 2012 (UTC)

  Looking strictly at what noun is grammatically modified by the adjective "free", you are right about "free occurrences", but these are still free occurrences of variables. About Wolfram Alpha, you are right that it is not exactly a dictionary, even though most of its entries seem to be restricted to definitions. But you know what? MWO seems to have "free variable": go to http://www.merriam-webster.com/dictionary/variable, scroll down in the list of terms at the top until you find "free variable", and click on it; I cannot view the entry though, as it requires complete access to the dictionary. MWO also has independent variable in the same scrollable list. --Dan Polansky (talk) 13:04, 11 November 2012 (UTC)

  Yes the occurrences are occurrences of variables, indicating that "free variable" is undoubtedly a concept, but if "free variable" were an independent lexical term, then it would always occur as "free variable" without any changes such as substituting the word variable for occurrence. --WikiTiki89 13:26, 11 November 2012 (UTC)

  Are you saying that if "look up" 'were an independent lexical term, then it would always occur as' "look up" rather than, say, "look it up"? By the way, I do not know what a "lexical term" is, so I will assume that it is the same thing as "term". However, there is no doubt that "free variable" is a term; the open thing is whether it is a sum-of-parts term, and whether Wiktionary users benefit from our having the term. Nor do I know what an "independent term" is, and what its opposite--dependent term--would be, and what it would be dependent on. --Dan Polansky (talk) 17:49, 11 November 2012 (UTC)

  I think the only way we could reasonably keep this entry is if we make it clear in the entry, and the entries of any other adjective + noun collocations, that the sense applies to any combination of (adjective) with something that is a (noun) even if (noun) is not part of the phrase. Which I think is rather overboard. After all, using that sentence we could enter free man and argue that "free slave" is really just that sense because a slave is a man. Or something like that. It's a slippery slope that I don't think we should get into. —CodeCat 13:31, 11 November 2012 (UTC)

  We should probably have an entry for free man per WT:COALMINE, per freeman and Google Ngram view search, although "free man" does not seem all that much more common than "freeman". We should have free market, free love, free speech, free energy, free will, and free software, as we have. Most or all of these terms are separable. I do not know what "free slave" is other than an oxymoron; it could be a recently freed slave, just that the term does not find much use in google books:"free slave", despite the high number of hits which turn out to be false positives upon inspection. --Dan Polansky (talk) 17:49, 11 November 2012 (UTC)

• Delete based on Ruakh's telling point. Whether there are other entries that suffer from the same defect is just a kind of makeweight argument, not a substantive one. We are in the unfortunate position of having to argue many of these cases from first principles. It's great for practicing argumentation but not so much for building or maintaining a dictionary, or short-attention-span encyclopedia, or phrasebook - whichever we are or want to be. DCDuring TALK 20:33, 11 November 2012 (UTC)

  I found the telling point of Ruakh's analysis to be that variables are not said to be free but to occur free, I would imagine because "x is free" is already overloaded with too many meanings. And in that case it's very clear that x is a variable, even though the word "variable" isn't explicitly used. So from that investigation I would also allow occur free. DAVilla 23:29, 11 November 2012 (UTC)

  But occur is a copula, so it can connect a subject with any predicate in theory. Variables can also occur bound, green, upside-down, cleared, in-scope etc. —CodeCat 23:43, 11 November 2012 (UTC)

  Re: "Whether there are other entries that suffer from the same defect is just a kind of makeweight argument, not a substantive one.": I don't think so. Listing other entries to which a principle applies helps evaluate the principle. The key question, one that you so often like to disregard, is whether the user of the dictionary is better off when the dictionary lists free variable, free market, free love, free speech, free energy, free will, and free software in separate entries rather than listing them at free. My point is that (a) it is not perfectly clear that "free variable" is a sum-of-parts term (a CFI consideration), and that (b) even if it is a sum-of-parts term in some technical sense, the dictionary is better off having the term (a non-CFI consideration). I supported (b) with (c) pointing out to Merriam-Webster Online having "free variable", and (d) my personal report of how I, a user of the term "free variable", happen to store the term in my mind. --Dan Polansky (talk) 18:36, 12 November 2012 (UTC)

• Delete per Ruakh.​—msh210℠ (talk) 07:22, 13 November 2012 (UTC) Redirect to free.​—msh210℠ (talk) 15:23, 13 November 2012 (UTC)
  • I mean, redirect to free preferably; otherwise, delete.​—msh210℠ (talk) 02:44, 6 February 2013 (UTC)
• Keep. A clear set term. In the OED, Chambers, and dozens of other dictionaries. Ƿidsiþ 07:52, 13 November 2012 (UTC)
• Keep, as it's not sum of parts. Our definition needs work, but If you'll note the linking Wikipedia article, a free variable is defined as a form of notation, and not as the expected abstraction. That is, a mathematical (deprecated template usage) variable could refer either the notation or the thing represented by the notation (our entry for variable has both senses). So, unless (a) the term free variable can apply to both, and (b) the term free can be used to qualify other mathematical terms, then I can't see us eliminating this entry. Further, the term (deprecated template usage) free has two definitions specific to mathematics, and so we have another pair of possible meanings. That's four possible definitions that might be assembled from senses specific to mathematics, but there is only one actual definition possible by my understanding. As a result, I feel we should keep this entry. --EncycloPetey (talk) 00:05, 5 January 2013 (UTC)

kept no obvious consensus for deletion -- Liliana • 11:00, 20 April 2013 (UTC)