Talk:horseshoe curve

RFD discussion: February–April 2021
NISoP: horseshoe ("The U shape of a horseshoe") + curve.

DCDuring (talk) 17:06, 27 February 2021 (UTC)


 * Keep &mdash; Dentonius 18:37, 27 February 2021 (UTC)
 * Keep. I can't see any sense in deleting this, nor any for deleting . DonnanZ (talk) 18:46, 27 February 2021 (UTC)
 * For reference, ngram data for {horseshoe, hairpin} × {curve, bend, turn}. Anyways, looking at the term, I never would have guessed it had a meaning specific to railways and roads (which google books seems to confirm). And there isn't a sense at curve that can recover this meaning (I don't count "A gentle bend, such as in a road." since it gives roads as an example rather than a requirement). Colin M (talk) 21:51, 27 February 2021 (UTC)
 * The meaning is not specific to railroads and other roads; it can also refer to the course of a river or the shape of a coastline. And a curve in a road does not have to be gentle. --Lambiam 19:33, 1 March 2021 (UTC)
 * Interesting - but are you sure those aren't just authors expressing the concept of a curve shaped like a horseshoe in a SoP way? In the way that one might talk about an "S curve", or a "hockey-stick curve", or whatever? In any case, the fact that over 90% of the gbooks results for that term relate to railways or roads still suggests there's a surprising degree of affinity with that semantic niche. Colin M (talk) 00:25, 2 March 2021 (UTC)
 * I am sure these uses are instances of authors expressing the concept of a curve shaped like a horseshoe in a totally SoP way. The issue here is whether our entry isn't likewise an SoP. I can find uses of blue expanse meaning, specifically, “a large body of water”, such as a sea. Such specific meanings do not make this collocation any less soppy, and an entry for, defining it as “An expanse having a blue colour, often a large body of water”, would be misplaced. A pretzel shape is a shape formed like that of a pretzel, and a horseshoe curve is a curve in the shape of a horseshoe. The facts that yoga practitioners may be found twisted in a pretzel shape, or that such horseshoe-shaped curves occur on roads, do not justify entries. --Lambiam 11:20, 2 March 2021 (UTC)
 * Ah, I see. Well I can agree that many uses of "blue expanse" to describe a large body of water would be consistent with SoP use, because, well, bodies of water comprise a large proportion of the things we're likely to encounter which are blue and expansive. But I would put it that, counterfactually, if >90% of uses of "blue expanse" were describing, say, regions of the United States that voted for the Democratic Party, then that would indicate an idiomatic usage. And I think the case here is closer to that. After all, if I'm just searching for an SoP description of something that is curved like a horsehoe, "U curve" would be equally appropriate, but if you search for that term on Google Books, only a tiny fraction are related to roads or railways. This is not consistent with the hypothesis that "horseshoe curve" is strongly associated with roads and railways because they simply represent a large fraction of entities that have such a shape. Colin M (talk) 21:41, 3 March 2021 (UTC)
 * Ah, I see, we were actually arguing the same point but in alternate realities. When above you wrote, “I never would have guessed it had a meaning specific to railways and roads”, I interpreted this as if you meant to imply that the term has a non-transparent meaning (one that one would not deduce from its parts) specific to railways and roads, thereby arguing that this is more than a sum of parts. But now I understand that you meant to say that in your opinion the term does not have a meaning specific to railways and roads, so your point is that the clause “often on a railway, where hairpin bends can't be used” in the present definition is not justified – a clause whose removal will definitely expose the essential SOP nature of the term. --Lambiam 13:41, 4 March 2021 (UTC)
 * No, your first guess was right (or close). I'm saying that 1) the evidence suggests there is a strong association between this term and the domain of roads and railways, 2) this association is surprising, and therefore 3) we should probably have an entry for this, since its meaning is more specific than just the SoP "any curve shaped like a horseshoe". As for the quoted part of the definition, the "often on a railway" text does seem justified (based on what I've seen in searches of Google Books etc.) I lack the domain knowledge to opine on the "where hairpin bends can't be used" portion. Colin M (talk) 03:40, 6 March 2021 (UTC)
 * A hairpin bend, like a hairpin, has a much sharper bend than a horseshoe curve, or a horseshoe bend in a river. Believe you me, driving uphill round a hairpin bend on a corrugated gravel surface isn't funny. You can't use a hairpin bend on a railway, but switchbacks and zigzags were used, where a train reverses. DonnanZ (talk) 22:56, 7 March 2021 (UTC)


 * I think there might just be enough in the "especially railways/roads where a very tight turn cannot be used" aspect to keep this, and I would also say that, while our definition presently reads "A curve in the shape of a horseshoe", in fact there seems to be more to it than that, since a "horseshoe curve" in the railways/roads context has inflections, while a horseshoe does not, or not typically. Mihia (talk) 21:49, 2 March 2021 (UTC)
 * But use of a term like horseshoe is always impressionistic. DCDuring (talk) 01:27, 4 March 2021 (UTC)
 * It's a proper noun too, so there's even less reason for deletion: . DonnanZ (talk) 22:18, 7 March 2021 (UTC)


 * Keep Purplebackpack89 14:33, 8 March 2021 (UTC)
 * RFD-kept. —Μετάknowledge discuss/deeds 22:34, 26 April 2021 (UTC)