Wikipedia:Reference desk/Archives/Language/2023 February 25

= February 25 =

Japanese text question
I saw this sticker on the window of a tram in Helsinki, Finland. The man in the picture is apparently saying something in Japanese. What does the text say? J I P &#124; Talk 01:16, 25 February 2023 (UTC)
 * "強制グレインダア 私の名前はアマカリスです よしクソ良い". It barely makes sense, and some kanji strokes are missing ("私" and "良"). Nardog (talk) 09:19, 25 February 2023 (UTC)
 * So does it have any meaning or is it just gibberish? J I P  &#124; Talk 10:20, 25 February 2023 (UTC)
 * Not exactly gibberish, I guess a forced translation would be something like "Forced grainder. My name is Amakarisu. Good fricking good." No idea what grainder is or what Amakarisu is a transliteration of. Nardog (talk) 12:40, 25 February 2023 (UTC)
 * It is a katakana rendering of the English word grinder, although the canonical rendering is グラインダー. In the context it may refer to chikan (frotteurism). --Lambiam 18:20, 25 February 2023 (UTC)
 * If it is then it's a mistake. Nardog (talk) 19:09, 25 February 2023 (UTC)
 * Random Googling implies that Amakariso is a term used in the Kinyarwanda language (Bantu family, spoken in Rwanda by 10m) and may mean capacity. Google also finds somebody called Amakaris on a site for social media, so it could, indeed, be a name.
 * Maybe a speaker familiar with Japanese Kanji / hiragana may shed some light on "forced grain". It could be a comment on the cuisine of the restaurant or some idiomatic expression.  --Cookatoo.ergo.ZooM (talk) 17:10, 25 February 2023 (UTC)

Introductory commas
In surveying my watchlist I occasionally come across edits like
 * In 1912, he stood for parliament . ..

where the user has inserted a non-parenthetical comma, and it annoys me but I don't know why. Fortunately I can usually revert the edit on some less specious grounds, but is there a term for the practice, is it a regional thing, and am I wrong? Doug butler (talk) 02:40, 25 February 2023 (UTC)


 * Not sure what style guides say, but I don't see anything wrong with the comma, since what precedes it specifies the circumstances in which the action of the main clause takes place. In a sentence beginning "As he perused the works of Catullus while sitting in his bathtub...", there would be a comma after "bathtub"... AnonMoos (talk) 03:43, 25 February 2023 (UTC)


 * The comma is clearly optional in such sentences, and if the sentence is longer and more complicated, then it may clarify how things are to be grouped. If everything is short and simple then I think it's best not to use the comma. --174.89.12.187 (talk) 05:18, 25 February 2023 (UTC)


 * It can also help, when the next word is a vowel, to make clear that the number is the year that something [the predicate] happened, rather than a count of the vowel noun that follows the numeral, as in "In 870 Saxons unified England." Confusion could also occur when the date is a title or part of a title, as in the 1619 Project or the musical 1776. —— Shakescene (talk) 06:36, 25 February 2023 (UTC)
 * A count of the vowel? --  Jack of Oz   [pleasantries]  07:59, 25 February 2023 (UTC)
 * Of course (being sleepy), I wrote "count of the vowel" when I meant "count of following noun" or (more pedantically) "count of the things, people or events represented by the noun that follows the numeral". One should never go on Wikipedia when sleepy, but that's a counsel of perfection that few of us follow. ;=) Now I'll be so bad as to go rewrite my words above. —— Shakescene (talk) 09:03, 25 February 2023 (UTC)


 * It looks very wrong to me. I think it's an American thing. DuncanHill (talk) 08:50, 25 February 2023 (UTC)
 * The Times then is guilty of a very un-British thing: "In 1912, Asquith's Liberal Government introduced its Irish Home Rule Bill." --Lambiam 17:27, 25 February 2023 (UTC)
 * The Times is not what it used to be. DuncanHill (talk) 17:48, 25 February 2023 (UTC)
 * That link is from 2007. Can you pinpoint just when the Times fell from grace? ←Baseball Bugs What's up, Doc? carrots→ 05:14, 26 February 2023 (UTC)
 * 1967. DuncanHill (talk) 09:57, 26 February 2023 (UTC)
 * So that was the year they first espoused comma-nism? ←Baseball Bugs What's up, Doc? carrots→ 15:53, 26 February 2023 (UTC)
 * At least establishing comma-nism with a human face, hopefully. It was a rough reign during the comma-nazis... 惑乱 Wakuran (talk) 21:42, 26 February 2023 (UTC)
 * human typeface you mean Fishing Publication (talk) 13:28, 27 February 2023 (UTC)
 * When I worked as a copyeditor (U.S.), the rule of thumb we used was that a simple introductory prepositional phrase, like "In 1912 ...", didn't usually need to be followed by a comma, whereas two or more prepositional phrases ("On the lower decks of the ship, ...") usually benefited from being followed by one. A comma after an introductory clause ("As he perused the works of Catullus, ..."), on the other hand, was almost always preferable. Deor (talk) 14:29, 25 February 2023 (UTC)
 * I use the rule of thumb that when I feel I should briefly pause when speaking the sentence, I introduce some punctuation at that spot. In the example sentence, that would be a comma. Copyeditors may not (and often do not) agree with my style; I have kind of given up challenging their changes unless a change introduces an ambiguity or an unwanted change in meaning. --Lambiam 17:37, 25 February 2023 (UTC)
 * Ah now, I would not pause between "1912" and "he stood". DuncanHill (talk) 17:48, 25 February 2023 (UTC)
 * Would it give you pause if a speaker did? --Lambiam 18:06, 25 February 2023 (UTC)
 * I might offer him my asthma inhaler. DuncanHill (talk) 01:24, 26 February 2023 (UTC)
 * https://separatedbyacommonlanguage.blogspot.com/2017/09/optional-commas.html Nardog (talk) 20:09, 25 February 2023 (UTC)
 * A good ref, thanks. Doug butler (talk) 23:30, 25 February 2023 (UTC)

Phonotactics question
Why onsets /ks/ and /ps/ are not possible in English? In Finnish they are possible, at least in loanwords. Although Finnish has far fewr word-initial consonant clusters than English, in this thing finnish is more allowable than English. --40bus (talk) 14:37, 25 February 2023 (UTC)


 * You really should have learned by now not to ask these senseless, inane "why" questions. How often do you need to be told: they are not answerable? Fut.Perf. ☼ 16:25, 25 February 2023‎ (UTC)
 * What kind of answer are you expecting, 40bus? Like almost all "Why" questions about language, an entire and complete answer to the question is "Because that is the way the language has happened to develop". ColinFine (talk) 16:27, 25 February 2023 (UTC)
 * Why does Finnish not have ejective consonants? --Lambiam 18:04, 25 February 2023 (UTC)


 * 40bus -- "S" always comes first in word-initial consonant clusters in English (spl, spr, str, skr, skw, sk, st, sl, sn, sf, sp, sm, and sw). In some languages (such as ancient Greek), /ts/ or /dz/ is kind of a single phoneme, and /ps/ and /ks/ occur in a wide range of phonological environments (they're some of the very few consonant clusters which can end an ancient Greek word), even if they're not single phonemes.  However, that's not the case in English... AnonMoos (talk) 23:26, 25 February 2023 (UTC)


 * IIRC, there is some variation on how much to nativise clusters in borrowings. I have /fθ/ in phthalate, /kθ/ in chthonic, and /tm/ in tmesis. And I also give Zwischenzug its German pronunciation even in English. But I have neither /pn/ in pneumatic nor /pt/ in pterodactyl. Also, doesn't /sf/ only appear in loanwords in English (e.g. sphere, sphenic, or for musicians sforzato)? Double sharp (talk) 01:00, 26 February 2023 (UTC)
 * I might sometimes pronounce the Greek letters ξ and ψ in English as /ksai/ and /psai/ respectively, especially when both are being used. (This pronunciation is mentioned on Wiktionary.) Double sharp (talk) 00:49, 26 February 2023 (UTC)
 * They may not have featured in natively English (i.e. West Germanic) words, but /ps/ is routinely used in modern English in Greek-derived words like Psionic, which as a science fiction fan I see and hear regularly.
 * No natural language is going to use all possible humanly reproducible sounds, because the necessary data can be encoded in far fewer, lessening the burden of learning and memory on its users. I believe there is a formal scientific treatment of this, but not being a professional linguist I can't recall it well enough to find a reference –, do you know of it? {The poster formerly known as 87.81.230.195} 51.198.55.125 (talk) 09:44, 26 February 2023 (UTC)


 * As a science-fiction fan I have also regularly seen the word "psionic" or "psionics", but I've never heard it. But since initial "ps" is regularly pronounced "s" in English, that's the natural way to pronounce "psionic", and it's the pronunciation given in Wiktionary and in [this dictionary]. If you think it's routinely pronounced with a P sound, you need to cite evidence. --174.89.12.187 (talk) 11:43, 26 February 2023 (UTC)
 * I don't have Reliable sources, but then again I'm not trying to insert this information into an article; I'm merely reporting what I hear (and say) in both general life (where also Psittacosis, for example, may be discussed) and at Science fiction conventions here in the UK. It might be that differences in what you and I hear reflect the way Classical Greek is taught academically on opposite sides of the Atlantic – I haven't studied it formally myself, but enough people in educated circles (into which SF fandom largely falls) in the UK have that their pronunciation of Greek-derived words may be a significant influence.
 * is a British dictionary, and it gives /ˌsɪtəˈkəʊsɪs/ as the BrE pronunciation. --82.166.199.42 (talk) 07:22, 27 February 2023 (UTC)
 * Then again, the 'p-' component of /ps/ may be sufficiently subtle that one doesn't notice it unless one is listening for it – humans have a tendency to hear what they expect rather than exactly what is annunciated, and sometimes literally can't hear certain sounds in (usually foreign) speech until they have been sufficiently exposed to and taught them: consider the differences between 𑀧 and 𑀨 (unbreathed and breathed p, as in "spin" and "pin"), and between 𑀩 and 𑀪 (unbreathed and breathed b) in Hindi, which English speakers generally don't notice even though the pairs exist in English, because in English the differences in each of them are not thought significant, so they don't have separate letters. {The poster formerly known as 87.81.230.195} 51.198.55.125 (talk) 14:07, 26 February 2023 (UTC)
 * Subtlety may be at play in < both  and, but not in many other cases, such as
 * < both (gnat ) and   (knee );
 * < both (pterion ) and  (ctenid );
 * < (bdelloid ).
 * --Lambiam 08:39, 27 February 2023 (UTC)
 * As I mentioned above, the two English words where I do realise initial Greek xi and psi as /ks/ and /ps/ are exactly the letter names xi and psi, because they would otherwise be ambiguous. That said, I have definitely heard xi rendered as /sai/ anyway, never mind the ambiguity. Double sharp (talk) 16:55, 27 February 2023 (UTC)
 * The only 'ps...' words listed in the Oxford English Reference Dictionary (because: single volume, within arms reach, and has pronunciations) where the initial 'p' is sounded are pshaw, psi and psst. -- Verbarson talkedits 21:30, 27 February 2023 (UTC)
 * To continue from Lambiam can;'t we include unvoiced fricative ?
 * Deprecated (to my knowledge) currency unit peseta /pəsɛtə/ ( ? )
 * or can't you think about for example /pt/ in pterodactyl ? A resource to consult for such pronunciation-spelling irregularities would be P for Pterodactyl  the recent book maybe you can find it online or a similar resource Fishing Publication (talk) 13:44, 27 February 2023 (UTC)
 * The entire point of that book is that the P is silent. It is actually called "the worst alphabet book ever" because the words don't illustrate the usual sounds of the letters. --User:Khajidha (talk) (contributions) 00:01, 28 February 2023 (UTC)

I wonder what other languages tend to simplify these clusters as English does. Looking first at Western Europe: in all the examples I can think of, French and German keep initial /ks/ and /ps/ intact in Greek loans. But according to English Wiktionary, Italian has both xilofono with /ks/ and silofono with /s/, and Spanish writes xilófono but pronounces /s/ instead of /ks/. Portuguese has xilofone with /ʃ/, which orthographically makes sense. Meanwhile Italian preserves /ps/ in psicologia but Spanish simplifies it to /s/ in psicología. English Wiktionary showcases a third option for Brazilian Portuguese psicologia: either preserving the cluster as /psi/, or breaking it up as /pisi/. I haven't looked much farther afield, because the Slavic languages are already cluster-happy, and because outside Europe, the borrowing has probably been filtered through some other European language on the way from Greek (certainly the case for Japanese サイコロジー). Double sharp (talk) 16:42, 27 February 2023 (UTC)
 * Summoning you for your knowledge. :D Double sharp (talk) 16:56, 27 February 2023 (UTC)
 * For a lot of languages, I suspect a less-assimilated word like 'pterosaur' will turn out to be like 'Dnieper' in English -- people may think they pronounce it like it's spelled, and dictionaries may even record it that way, but if you find recordings you'll discover that's not true. (Just as some English dictionaries falsely record 'Dnieper' with an initial ['dn], but their own sound file belies that claim, and similarly with African place and personal names that have an initial NC cluster.) Or people who are well-read may incorporate foreign consonant clusters into their speech but most people will not. I've found with Esperanto that some people try to take Zamenhof's diction of "one letter one sound" literally, and insist that one should not (and they personally do not) assimilate e.g. voicing of adjacent obstruents, and even upload sound files to prove it, but again the sound files prove them wrong. So if a dictionary does claim [ks] or [ps] clusters in these words for a language that does not otherwise have such initial clusters, I'd take that with a grain of salt; I suspect that if you recorded your pronunciation of 'tmesis' you might find an epenthetic schwa rather than an actual [tm]. (Though that might be complicated by aspiration.) — kwami (talk) 19:34, 27 February 2023 (UTC)
 * Both french and english wikt. claim that 'ptérosaure' is pronounced with a \pt\ [wikt-fr] or /pt/ [wikt-en] (I don't know what the difference is supposed to be), but when I check the sound files at Larousse, I only hear a [t]. They do have a clear [ps] in 'psychologie', but 'xylème' is 'gz' aand for 'xylophone' I hear just [z]. I came across an excellent online French dictionary recently that had transcriptions of most words, but unfortunately I forget where and all my stuff is in storage. — kwami (talk) 19:52, 27 February 2023 (UTC)
 * Having just repeated the word to myself several times, I think you're correct about my pronunciation of 'tmesis'. I think it's really more like [tʰə̥ˈmiːsɪs] with a quick and very devoiced schwa (which probably explains why I failed to notice it). This is also how I handle Slavic clusters when referring to Slavic place names; maybe without the devoicing, but certainly with a quick epenthetic schwa inserted. Then again, judging by this native pronunciation of Gdańsk, perhaps that's how you're meant to do it anyway in Polish.
 * I would guess then that most dictionaries are probably being overoptimistic indeed. Double sharp (talk) 20:08, 27 February 2023 (UTC)
 * After looking up Fr. 'xylème' and hearing a clear [gz] (amended above), I vaguely recall that being the prescribed pronunciation in French, so it appears there is a cluster, just not the Greek cluster.
 * Perhaps the [z] in 'xylophone' is because that word is more assimilated than 'xylème'. (Though it is a different speaker. Wikt-fr transcribes it \ɡzilɔfɔn\. If one of the native speakers chosen by Larousse does not use the prescribed pronunciation, maybe that says something about the reality of the situation.)
 * Many languages use non-phonemic epenthetic schwas or aspiration to break up consonant clusters. E.g. in the Pacific Northwest of N.Am., the obstruents in those 'vowelless' words all have audible releases. (Phonetically they're not challenging; the problem is trying to define what a 'syllable' is.) But East Slavic /dn/ is not like that: it's [dⁿ] with a nasal release, not a released [d] followed by an [n]. (I've heard that a lot recently with the war in Ukraine.) Also, we have something different with Brazilian Portuguese, where a phonemic vowel breaks up the cluster. That's the solution you would get with Japanese or Swahili. — kwami (talk) 20:27, 27 February 2023 (UTC)
 * Then, I suppose the situation must be that for more learned words that aren't used so often, we might have a cluster, but for words like xylophone which are common, the phonetic reality is simplification to the native phonological system.
 * My impression with Russian is that it's indeed less happy than Polish to insert epenthetic schwas. If my ears don't deceive me, another common cluster-breaking solution there seems to be making some of the consonants syllabic. So лгать sounds to me like [ɫ̩ɡatʲ] from the sound file, whereas Polish Lwów sounds rather more to me like [lǝvuf]. (It needn't be a voiced syllabic consonant: the English Wiktionary sound file for всплеск sounds to me like [fs̩plʲesk].)
 * But in Polish and Russian art songs they are set as single syllables anyway. It tends to sound to me like a bunch of acciaccaturas ("crushed notes"), e.g. Ewa Podleś singing Chopin (it's already obvious in the first line starting at 0:05).
 * P.S. I wonder how /sf/ ever became universal in English in sphere. It's not a native cluster, is it? Double sharp (talk) 20:44, 27 February 2023 (UTC)
 * I'm only guessing, but I suspect that some French speakers might simplify initial /gz/ to /z/ and maybe also /ps/ to /s/, and that one of Larouss's speakers does that and neglected to compensate with the recorded token. So that while the dictionary says they're pronounced /gz/ and /ps/, the first consonant is actually variable, and probably dependent on a number of sociolinguistic factors, only one of which would be how frequent the word is.
 * As for why /#sf/ is so well established in English, I suspect it's at least partly because we already had #sC clusters, along with spher-, sphinx and sphincter being relatively common words for it to become established with. We also have #ʃC clusters from Yiddish; again, I suspect they were easy to borrow because they were a simple adjustment from the native #sC pattern. #ps and #ks are not difficult to pronounce, but they don't follow from native patterns. We do have people who try to pronounce initial #ts, e.g. in tsunami and tsetse, despite #ts being harder to hear than #ks or #ps, but we don't have an orthographic tradition from Greek loans to base those words on, so people fall back on spelling pronunciations. You see something similar with e.g. #pn from people with limited literacy, and with kids first encountering a word like pneumatic when learning to read.
 * Again, only guessing, and someone here might actually know, but it might be that Greek-derived initial ps, ks, pt, bd, pn etc. once were pronounced that way in English, and that the simplification didn't occur in the borrowing process, but later through simple historical cluster reduction, and that for whatever reason (e.g. leveling with #sC, or simply a frequency effect that kept it robust), #sf wasn't reduced along with the rest. — kwami (talk) 23:43, 27 February 2023 (UTC)
 * I can personally testify that French people pronounce the "f" in Fnac and the "p" in "pneu" (or at least some of them did during the early 1980s). AnonMoos (talk) 23:32, 1 March 2023 (UTC)
 * P.S. This discussion has knocked loose a memory of a guy telling me in the early 1980s that you could tell pseudo-erudite people because they try to pronounce the word "pseudo" with with a "p" sound! -- AnonMoos (talk) 23:45, 3 March 2023 (UTC)

51.198.55.125 -- One main reason why no single language uses all possible phonemes, is that they would not be sufficiently perceptually distinct... AnonMoos (talk) 00:48, 28 February 2023 (UTC)