Talk:English language

Content moved from the phonology section

Regional variation in consonants

There are significant dialectal variations in the pronunciation of several consonants:

• The th sounds /θ/ and /ð/ are sometimes pronounced as /f/ and /v/ in Cockney, and as dental plosives (contrasting with the usual alveolar plosives) in some dialects of Irish English. In African American Vernacular English, /ð/ has is realized as [d] word initially, and as [v] syllable medially.
• In North American and Australian English, /t/ and /d/ are pronounced as an alveolar flap [ɾ] in many positions between vowels: thus words like latter and ladder /læɾər/ are pronounced in the same way. This sound change is called intervocalic alveolar flapping, and is a type of rhotacism. /t/ is often pronounced as a glottal stop [ʔ] (t-glottalization, a form of debuccalization) after vowels in British English, as in butter /ˈbʌʔə/, and in other dialects before a nasal, as in button /ˈbʌʔən/.
• In most dialects, the rhotic consonant /r/ is pronounced as an alveolar, postalveolar, or retroflex approximant [ɹ ɹ̠ ɻ], and often causes vowel changes or is elided (see below), but in Scottish it may be a flap or trill [ɾ r].
• In some cases, the palatal approximant or semivowel /j/, especially in the diphthong /juː/, is elided or causes consonant changes (yod-dropping and yod-coalescence).
  • Through yod-dropping, historical /j/ in the diphthong /juː/ is lost. In both RP and GA, yod-dropping happens in words like chew /ˈtʃuː/, and frequently in suit /ˈsuːt/, historically /ˈtʃju ˈsjuːt/. In words like tune, dew, new /ˈtjuːn ˈdjuː ˈnjuː/, RP keeps /j/, but GA drops it, so that these words have the vowels of too, do, and noon /ˈtuː ˈduː ˈnuːn/ in GA. A few conservative dialects like Welsh English have less yod-dropping than RP and GA, so that chews and choose /ˈtʃɪuz ˈtʃuːz/ are distinguished, and Norfolk English has more, so that beauty /ˈbjuːti/ is pronounced like booty /ˈbuːti/.
  • Through yod-coalescence, alveolar stops and fricatives /t d s z/ are palatalized and change to postalveolar affricates or fricatives /tʃ dʒ ʃ ʒ/ before historical /j/. In GA and traditional RP, this only happens in unstressed syllables, as in education, nature, and measure /ˌɛd͡ʒʊˈkeɪʃən ˈneɪt͡ʃər ˈmɛʒər/. In other dialects, such as modern RP or Australian, it happens in stressed syllables: thus due and dew are pronounced like Jew /ˈdʒuː/. In colloquial speech, it happens in phrases like did you? /dɪdʒuː/."

Regional variation

The pronunciation of some vowels varies between dialects:

• In conservative RP and in GA, the vowel of back is a near-open [æ], but in modern RP and some North American dialects it is open [a]. The vowel of words like bath is /æ/ in GA, but /ɑː/ in RP (trap–bath split). In some dialects, /æ/ sometimes or always changes to a long vowel or diphthong, like [æː] or [eə] (bad–lad split and /æ/ tensing): thus man /mæn/ is pronounced with a diphthong like [meən] in many North American dialects.
• The RP vowel /ɒ/ corresponds to /ɑ/ (father–bother merger) or /ɔ/ (lot–cloth split) in GA. Thus box is RP /bɒks/ but GA /bɑks/, while cloth is RP /klɒθ/ but GA /klɔθ/. Some North American dialects merge /ɔ/ with /ɑ/, except before /r/ (cot–caught merger).
• In Scottish, Irish and Northern English, and in some dialects of North American English, the diphthongs /eɪ/ and /əʊ/ (/oʊ/) are pronounced as monophthongs (monophthongization). Thus, day and no are pronounced as /ˈdeɪ ˈnəʊ/ in RP, but as [ˈdeː ˈnoː] or [ˈde ˈno] in other dialects.
• In North American English, the diphthongs /aɪ aʊ/ sometimes undergo a vowel shift called Canadian raising. This sound change affects the first element of the diphthong, and raises it from open [a], similar to the vowel of bra, to near-open [ʌ], similar to the vowel of but. Thus ice and out [ˈʌɪs ˈʌʊt] are pronounced with different vowels from eyes and loud [ˈaɪz ˈlaʊd]. Raising of /aɪ/ sometimes occurs in GA, but raising of /aʊ/ mainly occurs in Canadian English.

GA and RP vary in their pronunciation of historical /r/ after a vowel at the end of a syllable (in the syllable coda). GA is a rhotic dialect, meaning that it pronounces /r/ at the end of a syllable, but RP is non-rhotic, meaning that it loses /r/ in that position. English dialects are classified as rhotic or non-rhotic depending on whether they elide /r/ like RP or keep it like GA.

In GA, the combination of a vowel and the letter ⟨r⟩ is pronounced as an r-coloured vowel in nurse and butter [ˈnɝs ˈbʌtɚ], and as a vowel and an approximant in car and four [ˈkɑɹ ˈfɔɹ].

In RP, the combination of a vowel and ⟨r⟩ at the end of a syllable is pronounced in various different ways. When stressed, it was once pronounced as a centering diphthong ending in [ə], a sound change known as breaking or diphthongization, but nowadays is usually pronounced as a long vowel (compensatory lengthening). Thus nurse, car, four [ˈnɜːs ˈkɑː ˈfɔː] have long vowels, and car and four have the same vowels as bath and paw [ˈbɑːθ ˈpɔː]. An unstressed ⟨er⟩ is pronounced as a schwa, so that butter ends in the same vowel as comma: [ˈbʌtə ˈkɒmə].

Many vowel shifts only affect vowels before historical /r/, and in most cases they reduce the number of vowels that are distinguished before /r/:

• Several historically distinct vowels are reduced to /ɜ/ before /r/. In Scottish English, fern, fir, and fur [fɛrn fɪr fʌr] are pronounced differently and have the same vowels as bed, bid, and but, but in GA and RP they are all pronounced with the vowel of bird: /ˈfɝn ˈfɝ/, /ˈfɜːn ˈfɜː/ (fern–fir–fur merger). Similarly, the vowels of hurry and furry /ˈhʌri ˈfɜri/, cure and fir /ˈkjuːr ˈfɜr/ were historically distinct and are still distinct in RP, but are often merged in GA (hurry–furry and cure–fir mergers).
• Some sets of tense and lax or long and short vowels merge before /r/. Historically, nearer and mirror /ˈniːrər ˈmɪrər/; Mary, marry, and merry /ˈmɛɪɹi ˈmæri ˈmɛri/; hoarse and horse /ˈhoːrs ˈhɔrs/ were pronounced differently and had the same vowels as need and bid; bay, back, and bed; road and paw, but in some dialects their vowels have merged and are pronounced in the same way (mirror–nearer, Mary–marry–merry, and horse–hoarse mergers).
• In traditional GA and RP, poor /pʊr/ or /pʊə/ is pronounced differently from pour /pɔr/ or /pɔə/ and has the same vowel as good, but for many speakers in North America and southern England, poor is pronounced with the same vowel as pour (poor–pour merger).

Cumbersome Lead

The lead is showing signs of bloat as it keeps getting small additions. I am going to make some bold edits to the first paragraph now, but pre-empting queries as to why by opening this talk section.

1. I am removing lead discussion about where the Angles came from (too much detail).
2. Removing the bit about England being named for the area they migrated too as that is not quite right and needs a lot of unpacking (but consider, at least , that the Scottish lowlands also saw such migration).
3. The bit about closest languages is a fraught and debatable topic. I am avoiding it with new wording. Scots is English too, in the Old English period at least.

I may prune other parts soon (will see how this one goes first). Something I noticed in doing the edit is inconsistent referencing style. An editor has been inserting shorter footnotes (sfn) into the document that has not used these in the past. Do we have a view on the best reference style going forward? Sirfurboy🏄 (talk) 13:51, 7 September 2022 (UTC)

Not sfn, certainly. Cite-bandits should be strongly discouraged. Johnbod (talk) 14:01, 7 September 2022 (UTC)

Ah, I took a look at the referencing to see how big an issue it would be to sort it out, and noticed it is not as easy as I had supposed. There are over 300 sfn references in the page, and a lot more than <ref> style. Clearly someone had changed the whole page at some point so I took a look at page history and found some very significant activity in March-May 2015. All sfn references were converted between March and April of that year. The activity appears to have been to prepare the article for GA status. So I may have mis-stated above. It may be there is already an archived consensus for sfn. I will leave that alone for now. Sirfurboy🏄 (talk) 15:51, 7 September 2022 (UTC)

So looking at the second paragraph of the lead, my thoughts are that there is repetition with the first paragraph. We are again told of the West Germanic root, and Norse influence. That was already mentioned in the first paragraph. I don't like "mutated" and think "influenced" by Norse would be better. Sirfurboy🏄 (talk) 07:22, 15 September 2022 (UTC)

Non-Germanic Influence in English

"English is genealogically West Germanic, closest related to the Low Saxon and Frisian languages; however its vocabulary is also distinctively influenced by dialects of French (about 29% of modern English words) and Latin (also about 29%),"

I don't get it, aren't Low Saxon and Frisian (as well as all other surviving modern West Germanic) languages also distinctively influenced by French and Latin? I know modern German and Dutch, for sure, are heavily influenced by French and Latin and you can quite regularly see loanwords from these languages in these written languages. Also why are we still quoting this silly 29/29% figure for loanwords when that study was compiled in such a ridiculous way and included so many technical and rare synonyms that are rarely, if ever, used by the average English-speaker? Now it might not be quite AS heavy as in English, but Dutch and German certainly aren't THAT far off in terms of Latin and French influence. 2.99.93.88 (talk) 02:39, 11 September 2022 (UTC)

Please provide reliable published sources that can be consulted for discussion, otherwise there's really nothing to be discussed here in way of improving the article. BilCat (talk) 04:46, 11 September 2022 (UTC)

The thing is most other language articles seem to be using Swadesh lists in order to measure their word origins. If you did this with English, of course, you have a Germanic language with a significant minority of Latin loanwords (much like German or Dutch with around 25% foreign loanwords, usually from Latin or largely Latin languages like French, it might be around 30-35% in English)

Because of course all Swadesh lists do is take a set amount of the most commonly used words (usually around 1000) in languages and trace their origins.

This article seems to be citing an amateur study from decades ago where a guy basically went through thousands of business letters (which would be more formal and corporate anyway, and contain far more French and Latin synonyms and loanwords, as would formal German and Dutch business letters) and tabulated the words people used and traced their origins?

How is that a reliable source? That's not how linguists assess vocabulary at all. 2.99.93.88 (talk) 16:02, 11 September 2022 (UTC)

This "study" also doesn't seem to have distinguished between French words that come from Frankish (and ultimately Proto-Germanic), and French words which come from Latin. Around 30% of Old French vocabulary was of Frankish origin, and likely slightly higher still was the Germanic influence in Norman Old French) even modern French is still around 15% Frankish in origin (despite attempts to purify it of Frankish influence).

Many of the most commonly used French loanwords in English actually have a Germanic origin through Frankish (flag, banner, abandon, merchant, banner, standard, forest, common etc. etc. etc.)

Having said that, I don't know if it takes into account words from Old English that were of Latin origin either (Old English was around 25% Latin loanwords itself).

I feel like the article could elaborate more on this complex nature of language borrowing in European languages, since there is a common, erroneous perception that English is not a Germanic language due to many loanwords from other languages. 2.99.93.88 (talk) 16:16, 11 September 2022 (UTC)

Shouldn't be too difficult to add a bit of text that relaxes the hold that lexicon has on people who don't really understand how language works. Direct attention to structural components that actually characterize the language type -- morphology, syntax. Barefoot through the chollas (talk) 03:39, 15 September 2022 (UTC)

I think the sentence is okay (maybe we could change the percentages to something more vague like "almost a third" or "over a quarter" or "between a third and a quarter"). It describes the situation perfectly well, is clear that the French and Latin influence is in the number of vocabulary words, saying nothing of frequency of use, and when we get down to it, the discussion above is confusing West Germanic and Germanic. That is a problem because English did derive from West Germanic but many of those most frequent Germanic words in the language now come from Norse influence, which is North Germanic.

The one change I would like is to change "closest" back to "closely". I tried that change before, but it was changed back. I understand why we want to say English is closest to Frisian, but the concept of closest language is all so debatable. In a summary in the lead, I think it is enough that we single out Frisian and Low Saxon as the closely related West Germanic languages. Sirfurboy🏄 (talk) 07:07, 15 September 2022 (UTC)

L2 Speakers

As per Ethnologue, 2022 Edition, English now has over a billion L2 speakers , someone please update the article accordingly. Bodhiupasaka (talk) 20:29, 1 October 2022 (UTC)

a complement clause such as I saw him leave, where the main verb is to see, which is in a preterite form, and leave is in the infinitive

Seems unkind to tell naive readers that on the one hand saw is the preterite of to see, while leave is in the infinitive. I.e. the "bare infinitive" actually is the infinitive. Barefoot through the chollas (talk) 18:31, 23 December 2022 (UTC)

"29% of modern English words"

Does anyone know where the statistics about dialects of France/French being (about 29% of modern English words) and Latin influences (also about 29%) are ultimately coming from? I realize the same statistics appear on Foreign-language influences in English, which contains a cute lil pie chart. But that page also fails to cite sources. The chart itself cites a deadlink from AskOxford. List of English words of French origin gives 45% (also citing a deadlink). I like the idea of having percentages, but can we have some citations? Wolfdog (talk) 14:11, 27 December 2022 (UTC)

Good point, Wolfdog. Enthusiasm or lack thereof for percentages aside, has anyone serious ever really calculated the % of "modern English words" taken in from Old Norman, Old French or more recent French? A lexicon is not finite. A calculation, if not couched overtly as an approximate estimation (based on what?), would have parameters, limitations, along the lines of "x% of the [y number] most frequent words used in Modern English", along with identification of the corpus, such as print of various registers, oral recordings, etc. The results can be simplified -- espérons-le with minimal distortion -- for Wikipedia purposes, but the reference(s) should be here. Barefoot through the chollas (talk) 18:43, 27 December 2022 (UTC)