Fun guy?
« previous post | next post »
"A Conversation with Toby Kiers, World Champion of Fungus", NYT 1/14/2026 — lots of interesting science, and this bit of sociophonetics:
Q: How do you pronounce “fungi”? Is it “fun-guy,” or “fun-jai,” or “fun-jee,” or “fun-ghee”?
A: I’m all over the place. I’ll start a sentence with “fun-jee” and by the end I’ll say “fun-ghee.” There’s no wrong answer!
There are multiple pronunciations in current English use. More American dictionaries favour the pronunciation
/ˈfʌnd͡ʒaɪ/ or /ˈfʌŋɡaɪ/, while more British dictionaries favour the pronunciation /ˈfʌŋɡiː/ or /ˈfʌnd͡ʒiː/. However, all four pronunciations are in use in both countries.
FWIW, I checked a random sample of 62 "fungi" from the previously-mentioned NPR podcasts dataset, which contains 3,199,859 transcribed turns from 105,817 podcasts, comprising more than 10,648 hours. The tally was
| [ˈfʌnd͡ʒaɪ] | fun-jai | 44 |
| [ˈfʌŋɡaɪ] | fun-guy | 15 |
| [ˈfʌŋɡiː] | fun-ghee | 3 |
| [ˈfʌnd͡ʒiː] | fun-jee | 0 |
…which supports Wiktionary's summary, especially because at least one of the [ˈfʌŋɡiː] cases had a British accent:
There were also a couple of cases where the speaker didn't commit, e.g.
Update — or this case, where the interviewer tries to keep Americans from being confused by an Australian's pronunciation:
Andrew Read: One of the great things about this paper on the genetically modified fungi is it does raise the scientific interest a lot in the fungi.
Joe Palca: Fungi, or if you prefer, fungi or fungi or funguses, could be used to inject all manner of deadly toxins into disease-carrying mosquitoes.
Philip Taylor said,
January 25, 2026 @ 10:22 am
They are [ˈfʌŋɡiː] ("fun-ghee") for this Briton. Variations in the final vowel I can understand, but rendering of the "g" as [d͡ʒ] beggars belief — are the speakers totally unaware of the accepted pronunciation of "fungus" ?!
Rodger C said,
January 25, 2026 @ 10:25 am
Philip, no more than you're unaware of the accepted pronunciation of "g" before "e" and "i" in Latin derivatives.
Philip Taylor said,
January 25, 2026 @ 10:50 am
Ah well, as we were taught only Classical Latin at school (thank you, Mr Wolfson), I have to confess complete ignorance of any changes to pronunciation subsequent to that period …
Mark Liberman said,
January 25, 2026 @ 11:24 am
See the Wikipedia article on the Traditional English Pronunciation of Latin.
anhweol said,
January 25, 2026 @ 11:32 am
I tried to think of parallel examples, but could only come up with the much less common foci and loci (from focus and locus) where again apparently the soft pronunciation expected from the standard Latin borrowing rules competes with a hard pronunciation by analogy from the singular.
If one wanted to apply revived Classical Latin pronunciation fully to the English word, one would presumably have to modify the first vowel as well… (except in those parts of Northern England where the STRUT vowel comes out as [ʊ] anyway). But from at least Norman times to the nineteenth century, English speakers would have pronounced Latin systematically with soft c's and g's in the relevant positions.
Victor Mair said,
January 25, 2026 @ 11:39 am
Cf. /ˈfʌŋɡɪsaɪd/, /ˈfʌnd͡ʒɪsaɪd/.
Also /ˈfʌndʒɪbəl/ for "fungible" (adj. "able to be substituted for something of equal value or utility", which comes from a completely different root than "fungi; fungicide":
=====
1765 as noun, 1818 as adjective, from Medieval Latin fungibilis, from Latin fungor (“to perform, discharge a duty”) + -ible (“able to”). Originally a legal term,[1] going back to Roman law: res fungibiles (“replaceable things”).
https://en.wiktionary.org/wiki/fungible
TR said,
January 25, 2026 @ 12:16 pm
There's also magus / magi, where the affricate pronunciation in the plural seems more common. And if derivational relatives count, rancid / rancor.
Lolwhites said,
January 25, 2026 @ 12:17 pm
A mushroom goes into a bar. The barman says, "sorry, we don't serve mushrooms here."
"Why not?" asks the mushroom, "I'm a fungi."
Y said,
January 25, 2026 @ 2:49 pm
Is there a speaker bias in the NPR sample? For example, could most of the tokens have come from one interview with one mycologist, who happened to prefer "fun-jai"?
Mark Liberman said,
January 25, 2026 @ 3:58 pm
@Y — "s there a speaker bias in the NPR sample? For example, could most of the tokens have come from one interview with one mycologist, who happened to prefer "fun-jai"?"
For that reason, I didn't take more than 2 tokens from any single interview, and usually that was one from the interviewer and one from the interviewee. Since the instances were chosen at random from a somewhat larger set, I didn't need to intervene in that way very often.
Would the percentages be stable across (American) interview outlets, times, and places? Probably not; but I'd wager that "fun-jai" is going to come out on top all the same.
J.W. Brewer said,
January 25, 2026 @ 4:17 pm
I think I go with "fun guy," because for me (as with focus/foci, suggested above) the keep the consonant-the-same-as-in-the-singular intuition overwhelms the "Traditional-English-Pronunciation-of-Latin" pronunciation of "g" before "i." Obviously there are other lexemes in my idiolect where I change consontants in the plural, like wolf-to-wolves, but apparently the hard-g/soft-g variation doesn't work like that in my mouth.
Like Philip Taylor, I was only taught Latin in reconstructed-Classical pronunciation, but when e.g. we were learning our first-declension nouns and were taught that "vagina" (="sheath") was pronounced /wa.ˈɡi..na/, that did not lead us to change our prior understanding of the English pronunciation (i.e., /vəˈdʒaɪ.nə/) of the English noun of the same spelling derived therefrom.
Michael Watts said,
January 25, 2026 @ 8:37 pm
That's an interesting claim.
So, you're completely ignorant as to the English pronunciation of the English words general, regenerate, agile, and frigid?
And you're also ignorant of the differential pronunciation of "c" between all words ending in -ic ("electric" and all words ending in -icity ("electricity")?
As far as the writing system is concerned, there's not much ambiguity about how the "g" in fungi "should" be pronounced. It's true that pronouncing it with /g/ is common – I would have said that "fun guy" was the most common pronunciation – but all prescriptivism clearly pushes against it, not toward it.
Barbara Phillips Long said,
January 25, 2026 @ 9:02 pm
I am on team fun-jai, but I am also aware of a pronunciation divide concerning a related organism — lichen:
A lichen (/ˈlaɪkən/ LY-kən, UK also /ˈlɪtʃən/ LITCH-ən) is a hybrid colony of algae or cyanobacteria living symbiotically among filaments of multiple fungus species, along with bacteria[1][2] embedded in the cortex or "skin", in a mutualistic relationship.[3][4][5][6][7] Lichens are the lifeform that first brought the term symbiosis (as Symbiotismus[8]) into biological context.
https://en.wikipedia.org/wiki/Lichen#Reproduction_and_dispersal
Anyway, I pronounce the g in fungus as I do the g in “guess” or “gun.” I pronounce the g in fungi as I do the g in “giant,” not as I pronounce the g in “given.”
HS said,
January 25, 2026 @ 9:24 pm
To this New Zealander it's definitely fung-ghee, though I certainly wouldn't swear that all my fellow countrymen and countrywomen share my preference. (Note by the way the extra "g". That "ng" should of course be an "n" with a little tail but I can't be bothered with IPA. And as far as I can see that actually increases the number of possibilities from 4 to 8.)
Or at any rate, that's how I would pronounce it if I was reading the written word "fungi". But in speaking and writing I actually prefer "funguses". Or more precisely, I think I write "fungi" and say fung-ghee when I am referring to one of the kingdoms of life in a scientific context, and "funguses" when I am referring to the individual organisms that are killing my apple tree. (This preference extends to other classical plurals as well – I prefer standard English plurals where possible. I'm sick of people who tell me that the plural of "octopus" is "octopi", but even more sick of know-it-alls who then smugly correct them and tell me that it's actually "octopodes" because it comes from Greek. I knew that already, but I'm speaking English and it's "octopuses" as far as I'm concerned! And don't even get me started on the subject of people in New Zealand who try to tell me that Maori words borrowed into English should take zero plurals (the politically correct thing to do these days) because they have zero plurals in Maori. But that's a whole different subject and I digress….)
Roger Lustig said,
January 25, 2026 @ 9:50 pm
If we want to enforce the hard G, we should put an H after it, as the Italians do. We know how to pronounce "ghetto" and (shudder) "Ghislaine."
Michael Vnuk said,
January 26, 2026 @ 3:39 am
The plural of 'cactus', ie 'cacti', also shows variant pronunciations ('-tee', '-tie'). I wonder if there is any correlation in pronunciation of 'fungi' and 'cacti'. In my case, I was taught 'cac-tie', but I can't confidently remember what I was taught (Australia, 1960s and 1970s) for pronouncing 'fungi'. I think it was 'fun-ghee' rather than 'fun-guy'. Normal English plurals would help a lot.
Michael Watts said,
January 26, 2026 @ 4:26 am
Does it? There is no available contrast between /fʌŋ.ɡəs/ and /fʌn.ɡəs/ – they're both realized as [fʌŋ.ɡəs]. We could ask the same question about "fun guy", where a speaker might devote some effort to distinguishing the /n/, but there's a word break involved there. "Fungi" is always going to use [ŋ].
I'm still confused that Studio Ghibli chose a spelling that unambiguously indicates the wrong pronunciation of their name. (Which, in Japanese, is "Jibuli" ジブリ.)
Philip Taylor said,
January 26, 2026 @ 5:02 am
« So, you're completely ignorant as to the English pronunciation of the English words general, regenerate, agile, and frigid? And you're also ignorant of the differential pronunciation of "c" between all words ending in -ic ("electric" and all words ending in -icity ("electricity")? ». No, that was not what I said. I regard those words as English, and simply follow "standard practice" when speaking them.
But "fungi" is in a different category because it has a clear Latin plural marker ("-i", not "-ses") and therefore I follow Classical Latin pronunciation to the best of my ability. Whence [ˈfʌŋ·ɡiː], not [ˈfʌŋ·ɡaɪ], and definitely none of the [ˈfʌn·dʒ*] variants.
Incidentally, in re Victor's earlier mention of "fungicide", the LPD (CD-ROM edition) has [ˈfʌŋ ɡɪ saɪd] but the (female) speaker clearly says [ˈfʌn dʒɪ saɪd] .
Also, I don't think anyone is seeking to enforce the hard G — we are simply reporting how we speak, not seeking to dictate how others should speak …
Michael Vnuk said,
January 26, 2026 @ 5:44 am
So why does the article's heading have 'fungus' and not 'fungi'? If it was another group of organisms, would the headline writer have referred to the champion of bacterium or virus or plant or elephant or dragonfly?
Peter Cyrus said,
January 26, 2026 @ 6:03 am
In Italian, Catalan, Castilian, and Portuguese, this problem was resolved by changing the spelling of the consonant (in accordance with their own standard orthographic rules) to keep it "hard" despite the following "soft" vowel:
cerco -> cerchi
cerco -> cerqui
busco -> busque (both Portuguese and Castilian)
That's why it's fungo -> funghi in Italian.
Philip Taylor said,
January 26, 2026 @ 7:00 am
"[W]hy does the article's heading have 'fungus' and not 'fungi'?" — Maybe to avoid raising the (clearly contentious) issue of " how should one pronounce 'fungi' " ?!
J.W. Brewer said,
January 26, 2026 @ 9:04 am
As a further datapoint on my idiolect, I have the "soft" g in "fungicide." (Not that I have occasion to say "fungicide" aloud all that often …) Using the basic morpheme as one element in a compound apparently licenses or tolerates a change of consonant in a way the singular-to-plural shift does not.
J.W. Brewer said,
January 26, 2026 @ 9:15 am
Another angle on Michael Vnuk's question might be that for certain sorts of fungus (see what I did there?), the singular sometimes functions as a mass noun. Contrast mushrooms (easily countable because easily sorted into individuals by sight) with mold (um, not so much).
Indeed, tweaking the headline to say "champion of the fungus" makes it sound to my ear notably odd in a way that "champion of the elephant" or "of the dragonfly" would not.
Consider by contrast something like "champion of algae," where algae although taking plural agreement is in many respects like a mass noun because virtually no one ever speaks of a single alga and if for some reason you needed to you might grope for a singulative construction akin to "grain of sand" before you remembered (or learned …) "alga" as a possible word to use.
Jarek Weckwerth said,
January 26, 2026 @ 9:35 am
@ Michael Watts There is no available contrast between /fʌŋ.ɡəs/ and /fʌn.ɡəs/ – they're both realized as [fʌŋ.ɡəs]. We could ask the same question about "fun guy", where a speaker might devote some effort to distinguishing the /n/, but there's a word break involved there. — The phonemicity of /ŋ/ in English is easy to argue for, and if it's phonemic then in principle the contrast absolutely is available. A fan guard has /n/ as an available option, while fang guard does not. And in terms of "word breaks", that's a risky topic in English. You could spell these two jointly and the phonological analysis would be the same.
Philip Taylor said,
January 26, 2026 @ 10:34 am
Whilst I don't go quite as far as JWB with "fungicide" (I retain the hard "g"), I do allow the second vowel to mutate from /iː/ to /ɪ/ …
Philip Taylor said,
January 26, 2026 @ 10:39 am
Oh, and if I may, JWB, do you revert to the hard "g" in "fungicidal" ? To my ear, /ˌfʌŋ dʒɪ ˈsaɪd {}^{ə}l/ would sound very strange indeed, whilst in /ˈfʌŋ dʒɪ saɪd/ it sounds fine but I would not personally use it.
J.W. Brewer said,
January 26, 2026 @ 11:38 am
I can't say have ever uttered the word "fungicidal" (or heard it uttered in a way that registered) but I accept that Philip Taylor may be onto something with his intuition that the hard g sounds better there. Something to do with the stress pattern?
Pedro said,
January 26, 2026 @ 11:42 am
@Peter Cyrus: That's not quite the same because, for example, the Italian word fungo is not a Latin loanword but a native Italian word (although obviously derived from Latin like most Italian words) and it fact it doesn't really mean "fungus"; it means "mushroom" (which not coincidentally is a type of fungus. The plural is formed regularly, according to the usual Italian rules, and is spelt funghi because that's how it's pronounced (i.e. with a hard G).
So it's no help to look at how Italian chefs talk when discussing mushrooms. When Italian mycologists are discussing the phylogenetic kingdom Fungi, they presumably use the Latin loanword fungi as we do, and I'd be interested to know how they pronounce that. My guess is something like /ˈfund͡ʒi/, which would contrast with funghi /ˈfunɡi/.
Interestingly, fungo and funghi are not, strictly speaking, derived from the Latin nominatives fungus and fungi but from the accusatives fungum and fungos, both of which would have had a hard G in Vulgar Latin, which presumably explains the hard G in funghi. (Actually I know that applies to French & Spanish but now I'm wondering where the Italian plural ending -i came from…it looks more like the nominative plural so maybe I've got that wrong.)
Philip Taylor said,
January 26, 2026 @ 1:01 pm
"I can't say have ever uttered the word "fungicidal" (or heard it uttered in a way that registered)" — those of us with fungal infections the our toe nails have perhaps used the word more frequently than most …
Mark P said,
January 26, 2026 @ 2:25 pm
On a slightly different note, is the use of “agendum” for a list of topics to be discussed common? I have seen it precisely once, coming from a fairly pedantic fellow.
Michael Watts said,
January 26, 2026 @ 3:55 pm
I don't think this is true. Yes, it's easy to show that English has a phonemic /ŋ/. But it's not true that the existence of a phoneme in some contexts means it is available to contrast with every other phoneme in every context.
For example, there is no contrast available in English between the TRAP, FACE, or DRESS vowels before /ŋ/. It's easy to show that those vowels can contrast in other contexts, such as before /k/ or before /m/, but this doesn't make it any more possible to contrast them before /ŋ/. Not only are there no words that might putatively exhibit such a contrast (not true of /g/, where we have beg/bag or leg/lag), speakers will struggle if you ask them to draw the contrast in invented words.
For nasals specifically, I think there is also a general problem that listeners aren't able to distinguish different nasals even when their articulation has been preserved. I think that English speakers, because they recognize the morpheme "hand", are likely to maintain an alveolar articulation when pronouncing the nasal in "handbag". But if the /d/ is not given audible release, the nasal will sound like an /m/ even though the alveolar articulation is present. Labial articulation is also present because of the /b/.
Fungicide appears in an insult in The Curse of Monkey Island, where it is pronounced with /g/. I would be surprised by a change in the pronunciation of "g" between fungicide and fungicidal, but someone did attest to that above, I guess…
There are some other words that seem relevant to the pronunciation of fungi. The same question arises when focus and locus are given Latin-style plurals; my sense is that these are likely to be pronounced with /k/.
I wouldn't agree that these are seen as Latin plurals in any case. Nobody knows Latin. People are doing their best to apply an English rule of unclear application, as documented here for orifice / orifi, danish / dani, apprentice / apprenti, and ridiculous [things] / ridiculi. Note that in that list of examples, "orifi" and "ridiculi" occur in a humorous context, but "dani" and "apprenti" do not.
This also goes the other way; it's not all that unusual to hear people generalizing from the plural vertices to the singular /vɚtəsi/.
Michael Watts said,
January 26, 2026 @ 3:58 pm
Another comparison that feels noteworthy is fetus; my sense is that the plural form feti is extremely marked and English speakers are likely to seek to avoid it. (This is also my sense for fungi – there are "fun guy" jokes, but the plural form is not clear and if you have another option you'll go for that.)
Philip Taylor said,
January 26, 2026 @ 5:13 pm
"Not only are there no words that might putatively exhibit such a contrast (not true of /g/, where we have beg/bag or leg/lag), speakers will struggle if you ask them to draw the contrast in invented words" — I agree with the first part, but not with the latter. For me, an L1 speaker of British English, I would have no difficulty in recognising (or producing) for example "sang" v. "seng", "lang" v. "leng", etc. The FACE vowel is, however, another matter.
Philip Taylor said,
January 26, 2026 @ 5:18 pm
"there are "fun guy" jokes, but the plural form is not clear and if you have another option you'll go for that" — again, I disagree. The only other option of which I am aware is "funguses" and I cannot imagine any situation in which I would use that form other than, for example, a quiz in which the question was " give a synonym for 'fungi' ".
Michael Watts said,
January 26, 2026 @ 5:28 pm
Note that the American host's attempt to rescue his listeners from /fʌŋɡi/ included "funguses", which makes sense because that is how the normal English plural would form. But the "other option" I was actually referring to is simply phrasing your sentence differently, and that option is available in almost all cases.
It's not necessary that all theoretical forms of every word exist, and the plural of fungus does not exist as clearly as, say, the plural of computer. Fungus isn't clearly defective as to the plural form either, but I think people will be uncomfortable if required to produce a plural, and look for something else they can say.
Philip Taylor said,
January 26, 2026 @ 5:32 pm
Well, next time I'm in the hotel I will try a simple test and ask each member of staff what the plural of "fungus" is. The sample will include native speakers of English, French, Vietnamese, Tagalog, Hindi and maybe one or two more that I have overlooked …
Jerry Packard said,
January 26, 2026 @ 6:12 pm
Of all the different tenni, I like table tennis best.
HS said,
January 26, 2026 @ 6:34 pm
Does it? There is no available contrast between /fʌŋ.ɡəs/ and /fʌn.ɡəs/ – they're both realized as [fʌŋ.ɡəs]. We could ask the same question about "fun guy", where a speaker might devote some effort to distinguishing the /n/, but there's a word break involved there. "Fungi" is always going to use [ŋ].
I did wonder about that very point. I tested it on myself by repeating all the various combinations over multiple times aloud and decided they were all clearly distinguishable. But admittedly I was pronouncing them reasonably slowly and carefully and the contrast may well disappear at normal speed.
For example, there is no contrast available in English between the TRAP, FACE, or DRESS vowels before /ŋ/. It's easy to show that those vowels can contrast in other contexts, such as before /k/ or before /m/, but this doesn't make it any more possible to contrast them before /ŋ/. Not only are there no words that might putatively exhibit such a contrast (not true of /g/, where we have beg/bag or leg/lag), speakers will struggle if you ask them to draw the contrast in invented words.
Presumably this reflects the difference between (your variety of) American (I presume) English and (my variety of ) New Zealand English, but the invented word pairs beng/bang and leng/lang are absolutely clearly and obviously distinguishable for me given my normal pronunciation of the TRAP and DRESS vowels. (I might note as possibly relevant here that New Zealand English has been subject to a systematic vowel shift and all the short front vowels have effectively moved up a step.)
Condign Harbinger said,
January 27, 2026 @ 5:10 am
To me, the more interesting aspect is the way foreign plurals are assimilated into English – or not, as the case may be. Nobody has problems with 'bonus' or 'rumpus'. Bankers never get 'boni'. Everybody has problems with 'octopuses', 'octopi', 'octopodes'. In 50+ years of electronics and computing, I can't recall anyone talking about a 'datum' as singular of 'data', perhaps because one on its own is a bit useless. On the other hand, as a reference point it's definitely a datum, and you usually only have one of those – the meaning has undergone binary fission.
So what are the rules about naturalisation, why 'omnibuses', 'Pyrocanthas', 'dilemmas' (and not even a revolting pedant would refer to 'dilemmae' or 'dilemmai'), but 'antennae'?
Oh, fun guy as UK English speaker, and I'm always very grateful to those Italian h's as insulators.
ajay said,
January 27, 2026 @ 6:29 am
"I'm still confused that Studio Ghibli chose a spelling that unambiguously indicates the wrong pronunciation of their name. (Which, in Japanese, is "Jibuli" ジブリ.)"
In this case the Japanese pronunciation and spelling are wrong and the English spelling is correct, and I am surprised by this because the founder of the studio should have known how to pronounce it properly. The studio was named, by its founder, after this aircraft https://en.wikipedia.org/wiki/Caproni_Ca.309 the Caproni Ca.309 Ghibli – Ghibli, pronounced with a hard G (hence the h), is an Italian word meaning a hot desert wind, derived from the Arabic.
ajay said,
January 27, 2026 @ 6:32 am
Bankers never get 'boni'.
Though they do get maluses – a monetary penalty applied to a deferred bonus. https://www.bankofengland.co.uk/prudential-regulation/publication/2015/pra-expectations-regarding-the-application-of-malus-to-variable-remuneration-ss
ajay said,
January 27, 2026 @ 6:37 am
I might note as possibly relevant here that New Zealand English has been subject to a systematic vowel shift and all the short front vowels have effectively moved up a step.
Brought home to a friend of mine very forcibly when he was trying to find his connecting flight to Wellington in a state of considerable jetlag, and the nice lady at the counter said "Wellington? You'll need to go to the Mystic Chicken."
"…wha?"
"The Mystic Chicken. If you want to fly to Wellington, that's the Mystic Chicken."
"The mystic…"
"The Mystic Chicken, yes. It's over that way."
And this went on for rather longer than it should have, and the poor chap had developed quite a clear mental picture of the chicken in question – a large beast, he thought, like one of the Eagles from "Lord of the Rings", with a few huddled passengers on its mighty back and an air of calm wisdom – before he realised that he was being directed towards Domestic Checkin.
Philip Taylor said,
January 27, 2026 @ 8:23 am
I love it, Ajay. Sadly I no longer fly anywhere near as much as I once did (my wife no longer works for Etihad, which explains everything) but the next time I do so I will make sure to ask at the Enquiry desk where Mystic Chicken is !
Rodger C said,
January 27, 2026 @ 10:52 am
Another comparison that feels noteworthy is fetus; my sense is that the plural form feti is extremely marked and English speakers are likely to seek to avoid it.
That's because foetus is a 4th-declension u-stem, and there's no word *foeti
Michael Watts said,
January 27, 2026 @ 11:26 am
I don't think a pedant is the personality type you're looking for here. A pedant can't refer to 'dilemmae' because 'dilemma' is not a first declension noun and its plural form is 'dilemmata' (compare stigma / stigmata). Omnibus has a similar problem; it's already plural and doesn't really function as the name of an object at all, not being in the nominative case. Pyracantha has no such issues.
But as a matter of usage this has to be restricted to a historical argument. Relatedly:
This can't actually explain the phenomenon, because it is not known to the people avoiding the form. The best-known (and clearly unmarked) example of the English plural marker is "octopus / octopi", which is also invalid in the source language.
Jarek Weckwerth said,
January 27, 2026 @ 1:52 pm
Michael Watts Yes, it's easy to show that English has a phonemic /ŋ/. But it's not true that the existence of a phoneme in some contexts means it is available to contrast with every other phoneme in every context. — True, but my other examples demonstrated that /ŋ/ vs /n/ is available in very similar contexts, and if you want a 100% identical vocalic context, then sun gasometer vs. sung gasometer.
I think there is also a general problem that listeners aren't able to distinguish different nasals even when their articulation has been preserved — That is far too wide-reaching.
Regarding the labial assimilation, the elision of /d/ is not sufficient, since both nb and mb are possible. The labial closure must at least overlap with the alveolar closure (if that is in fact left behind) into the nasal.
HS said,
January 27, 2026 @ 5:05 pm
@ajay,
The Mystic Chicken is an old joke and it's appeared on Language Log before, but I still love it! I think some enterprising Kiwis were even planning on setting up a fried chicken joint called Da Mystic Chicken at some point…
HS said,
January 27, 2026 @ 5:25 pm
"fried chicken joint" was perhaps an unfortunate choice of words there!
HS said,
January 27, 2026 @ 9:22 pm
I realise now that my comment above that there were really 8 possibilities, not 4, was not really thought through properly. I was thinking purely in terms of mathematical combinations and not phonological properties. (Also, it was really just a quick, last-minute addition to my broader comment, spurred on by the fact that Professor Lieberman, following the New York Times, refers to my preferred plural form as "fun-ghee" whereas I think it should be referred to as "fung-ghee".) Thinking it through a bit more carefully I realise what I wrote doesn't really make sense. The singular fungus is pronounced fung-gus (i.e. it has an /ŋ/, or at least an [ŋ], for everybody I presume) so I think you'd expect people to use a plural that has an [ŋ], unless there is some phonological constraint that determines otherwise and results in a plain [n] (which seems to be the case with the fun-jai and fun-jee plurals given by Professor Lieberman, even though I personally find fung-jai and fung-jee both perfectly pronounceable and acceptable.) So not all 8 of the theoretical mathematical combinations are actually going to occur as I suggested.
I think the more interesting question is whether the singular "fungus" (i.e. fung-gus with a /ŋ/, or at least an [ŋ]) "had to be" that way at all, or whether that [ŋ] is essentially "accidental" and in a different "possible world" (to borrow a term from philosophy) it could just as easily have been fun-gus with an [n] instead of an [ŋ]. Michael Watts I think would say no, it is phonologically constrained to be fung-gus with an [ŋ], because an /n/ would be realised as an [ŋ], but for me in my variety of English I'm not so sure. I find "fun-gus" perfectly acceptable and pronounceable and I don't think I would somehow be constrained to pronounce an underlying /n/ as an [ŋ]. (And elsewhere Michael Watts says that speakers will struggle to produce and distinguish between an invented pair of words like "beng" and "bang", but as I noted above, I find it trivially easy to produce both these words and they are quite clearly distinct for me – and Phillip Taylor says exactly the same thing about British English. So clearly Michael Watt's variety of English differs from my variety of English in important and relevant ways here.)
So for me as a speaker of New Zealand English I think the contrast between fung-gus and fun-gus is perfectly possible in principle and it's essentially just accidental that in this world it's fung-gus and not fun-gus. So if we consider all possible worlds then I think all 8 combinations for the plural of "fungus" are actually possible, as I originally said! (I'm ignoring any historical or etymological considerations here and just considering the current situation for me today as a speaker of New Zealand English.) But of course, since we're only in this world, any actual corpus search is not going to turn up all those 8 possibilities (so I can never be proved wrong!).
Having said that, I'm finding it hard in practice to come up with any words that are similar to fun-gus in pronunciation (though there is "vanguard" which is fairly similar), so Michael Watts seems to be right in practice, at least in this actual world. But I don't think he's right in other possible worlds.
(For the philosophically inclined, there is also the curly* question of how different another world can be from this one and still be considered a relevant "possible world".)
*Wictionary records this as "(Australia) Complicated and difficult; knotty."
HS said,
January 27, 2026 @ 9:28 pm
Oops, I wrote Lieberman instead of Liberman. My apologies to Professor Liberman.
Bob Ladd said,
January 28, 2026 @ 1:48 am
Bankers never get 'boni'.
They do in German.
Michael Watts said,
January 28, 2026 @ 2:56 am
I think you're going way too far there. You could, with effort, draw a distinction in your personal speech. I tend to suspect that you'd need to spend more time on those syllables than a natural rate of speaking would entail, and you'd draw some bemused comments. But it can be done and sometimes it is done.
But as time passes, it's guaranteed that people won't make that effort and you'll see assimilation. It isn't accidental that it's /fʌŋɡəs/ today – the concordance between /ŋ/ and /g/ is necessary if you take into account the fact that the word is not brand new. [fʌnɡəs] isn't stable enough to persist.
Michael Watts said,
January 28, 2026 @ 3:25 am
Well, I intended that statement to be scoped by the example that follows it. I was not hoping to claim that listeners are, in general, unable to distinguish different nasals. But I am claiming that they're unable to distinguish different nasals some of the time.
I agree that to get the sound of an /m/ (even where technically you're producing a strange composite sound), you need to have the labial closure overlap the alveolar closure. But I don't think that would be a surprising way to produce "handbag".
My gut says that people probably produce "NBA" in a way that is audibly different from "MBA", but I'm not basing that on any kind of evidence. Feels relevant here.
DDeden said,
January 28, 2026 @ 12:30 pm
Pedro "…the Italian word fungo is not a Latin loanword but a native Italian word (although obviously derived from Latin like most Italian words) and it fact it doesn't really mean "fungus"; it means "mushroom" (which not coincidentally is a type of fungus…"
A minor technical point, in US English biology, a mushroom is only the above-ground temporary fruiting body (cap & stem) of a fungus, excluding the underground mycelium.
In general usage, though, your phrase is correct.
Philip Taylor said,
January 28, 2026 @ 12:32 pm
"The singular fungus is pronounced fung-gus (i.e. it has an /ŋ/, or at least an [ŋ], for everybody I presume)" — untrue. For me, "fungus" is /ˈfʌn ɡəs/ but "fungal" is /fʌŋ ɡ${}^{ə}l/. Don't ask me why, I have no idea, I am just reporting what I experience.
HS said,
January 28, 2026 @ 7:06 pm
But as time passes, it's guaranteed that people won't make that effort and you'll see assimilation. It isn't accidental that it's /fʌŋɡəs/ today – the concordance between /ŋ/ and /g/ is necessary if you take into account the fact that the word is not brand new. [fʌnɡəs] isn't stable enough to persist.
I think you're almost certainly right. I was just considering the inventory of English phonemes today, and explicitly said I was ignoring any historical or etymological considerations. fun-gus seems to me to be a perfectly possible English word today, but as you say, it probably wouldn't persist over time. (So putting it in my philosophical context of possible worlds, it depends upon what exactly you consider to be the relevant possible worlds, and that will depend upon exactly what question in the actual real world you are trying to address.)
HS said,
January 28, 2026 @ 8:45 pm
Having agreed with your second paragraph, I actually respectfully disagree with your first paragraph where you say "You could, with effort, draw a distinction in your personal speech." I find it easy to say fun-gus and to distinguish it from fung-gus, just as I find it perfectly easy to pronounce a word like "vanguard" as van-guard and not vang-guard. And Phillip Taylor provides an example of someone who actually pronounces fungus as fun-gus, according to his comment. So again, I think this reflects a difference between your variety of English and my and Phillip's varieties of English. And certainly your earlier statement that "[t]here is no available contrast between /fʌŋ.ɡəs/ and /fʌn.ɡəs/ – they're both realized as [fʌŋ.ɡəs]" seems to me to be simply false. It's true that saying fung-gus is easier that saying fun-gus, so I agree with your second paragraph that over time the pronunciation of the whole speech community would evolve to fung-gus, but that's a different claim from saying that it would require effort to draw a distinction, and certainly a different claim from saying that there is no available distinction.
Michael Watts said,
January 30, 2026 @ 5:47 pm
These are good points; a lot of what I have to say will echo some things I've said upthread.
For "vanguard", I agree that it's easy to draw a clear distinction. I have some misgivings about how much of that distinction is due to the nasal consonant – I think the vowel in "van" [rhymes with man] is meaningfully distinct from the vowel in "vang" [rhymes with sang], which might make it easy to draw a van-guard / vang-guard distinction without being able to draw much of a distinction in the [n] / [ŋ].
For self-perception of produced [n] vs produced [ŋ], I worry about something that was briefly mentioned above. You can easily perceive whether you're making the alveolar closure appropriate to [n]. But to the extent that that gesture is coarticulated with with a labial or velar closure, a listener may hear [m] or [ŋ] even though the alveolar closure is also present. This has implications for whether the contrast is "available" in the language. If you can easily produce the difference, but your audience can't easily receive it, the contrast is not "available" in most useful senses.
These two claims can be said to be different, or not. Let's suppose that 30% of people naturally speak in a way that causes them to draw an internal distinction between [-ng-] and [-ŋg-]: if you put them in an MRI, you can see the difference. The other 70% of people obey a personal allophonic rule that /-ng-/ often or always becomes [-ŋg-]. I think it's fair to say that, given this state of affairs, the distinction isn't available in the language.
For a similar example, there is a [real] allophonic rule that allows /-sj-/ to become [-ʃ-] in almost all circumstances. (Though I'd be surprised if it happened in "miss you".) The rule is not mandatory, and is not always applied even where it would be unsurprising. But I offer that the existence of this rule means that a hypothetical contrast between two words that differ only by an internal /-sj-/ vs an internal /-ʃ-/ is not available in varieties of English that have it. Most of the time, those words will be indistinguishable, and this is without the possibility that listeners will be literally unable to hear the difference.
Andrew Usher said,
January 31, 2026 @ 12:38 am
I think the distinction is potentially perceptible, as I've heard conspicuously non-assimilated versions of words like 'concrete'. Perhaps not all people can make them audibly distinct or the audible distinction is lost in all but slow/careful/forcefully-articulated speech, and certainly agree that a minimal pair based on something like /ng/ vs. /ŋg/ is not likely.
I don't think the quality of the vowel is essential at all here; I can alternate between van-guard and vang-guard with no change to the vowel (I normally use the latter – word-internal nasal assimilation is obligatory to me, and I'm not going to force myself to make exceptions), though I know most people would make a corresponding vowel distinction is natural speech. And the vowel in my example 'concrete' has much less differentiation available than does TRAP.
Jason M said,
January 31, 2026 @ 3:10 pm
The Andrew Read in the audio clip is not a native Australian English speaker but a Kiwi, which I think you can hear in his “e” vowel shift, but even if his accent has Americanized somewhat over the years, his undergraduate degree is from New Zealand, and he sounds pretty obviously from there when I have talked to him, at least to my American, non-expert ear. (Though non-expert, I can always tell where the captain of my airplane is from if she says she is talking from the “flight dick”!)
Another note: those of us who culture cells in dishes in a lab use “fungicide” on a daily basis and “fungicidal” not infrequently. In America, both are usually soft “g”, I thunk, though I will pay more attention now. I think many of us would say funGUY but funJicide fwiw.
Michael Watts said,
February 1, 2026 @ 8:40 pm
On the subject of nasal assimilation, I was led to be curious about the name of Baden-Württemberg.
My (AmE) perception of a native pronunciation is something like /bɑdn̩ ˈvuʔn̩ˌbɚg/. At least, this is what I would say in my best attempt to imitate the German. I might go for /y/ instead of /u/ with varying degrees of success, and I similarly might try to fade out the rhoticity a bit.
What specifically interested me is that, despite the spelling with "m", I quickly determined that if I was going to try to pronounce this, I would have to use /n/, matching the /t/, and I cannot produce /m/ to match the /b/, even if I want to.
Tom Dawkes said,
February 5, 2026 @ 10:29 am
I think we should say 'funguses'. We don't talk about 'muses' but 'museums'
Tom Dawkes said,
February 5, 2026 @ 10:30 am
I meant 'musea' not muses
Philip Taylor said,
February 5, 2026 @ 11:28 am
" We don't talk about 'muse[a]' but 'museums' " — ooh, I think I might from now on, Tom, as you have very kindly drawn my attention to this sloppiness on my part …