User talk:Kisseran/Proto-Indo-European phonology

Centum and satem

@ThaesOfereode This is a really minor thing to be pointing out, but I don't think lang templates or italics should be used for the terms centum and satem. I've just checked four introductory books: Clackson, Sihler, Szemerényi, and Beekes, and only Clackson uses italics for them. Also, MOS:FOREIGN says A rule of thumb is to not italicize words that appear unitalicized in major general-purpose English dictionaries. OED, Merriam-Webster, and Collins all list centum and satem unitalicized. Kisseran (talk) 11:44, 9 July 2026 (UTC)

Yeah, this was originally my argument when I took Weise's law to FAC. I'll change it back; I was just following what I thought might have been a new convention. ThaesOfereode (talk) 12:08, 9 July 2026 (UTC)

Accent → Morphophonology section?

@Kisseran: Thinking we could do something better here. What if we made the current accent section a morphophonology section then subordinated the edge cases here (ablaut, vrddhi, accent mobility, etc.), sum them up shortly, and put a bunch of {{main}}, {{further}}, or {{broader}} tags as necessary. This would help us argue that the topic is comprehensive at GT because they're overlapping and addressed on the main page, but are morphological to such an extent that they should belong in a morphology GT. What do you think? ThaesOfereode (talk) 15:09, 11 July 2026 (UTC)

I think this is an good practical compromise. I'd initially started out without a morphophonology section because I, like many of the prominent generative phonologists we are going to cite in those sections, don't really think that morphophonology is a real separate thing. I also thought that in principle, accent was not inherently morphophonological and if anything would fit into a prosody section more than anything. Having a prosody section is desirable in that it's a common section to have in phonology articles and so would be an intuitive arrangement to some potential readers. The downside to having a prosody section is that the usual conception of what prosody is doesn't really line up with what is broadly dealt with in prosodic phonology, and it doesn't line up precisely in the areas that we already have trouble sorting out. In any case, I think it might be easier to leave the thinking about section titles for after some of the prose is already written. Kisseran (talk) 15:23, 11 July 2026 (UTC)

Yes, probably wise. No need to pull the trigger yet but I wanted to get this thought out before I forgot it later on. ThaesOfereode (talk) 16:36, 11 July 2026 (UTC)

Consonant table

@ThaesOfereode I've been thinking about the consonant table and looking at the various handbook treatments of it. To be honest, I'm not sure that it's worthwhile to make a table with complete rows and columns for the entire consonant inventory. If you look close enough, none of the handbook treatments but Byrd actually have a big consonant table for all of the consonants. The stops are always presented in table form, one way or another, but it is really not neither useful nor practical nor aesthetically pleasing to sort the rest of the consonants by place of articulation. I've mostly copied over Byrd's presentation of the consonant table, with the exception of being more vague about the place of the laryngeals. I don't know how you feel about balancing Wikipedia genre conventions and servicing the quickly skimming reader versus academic conventions and avoiding unwanted implications. The other concern is that keeping stops in a table and not ordering the other consonants by column can look decent in more LaTeX-esque plain table formatting, but probably wouldn't look very good or coherent with Wikitable formatting.

Some extra comments about terminology whether or not we decide to keep the big consonant table:

• Wikilinking laryngeal is definitely misleading. Wikilinking palatal is arguably misleading too. The others could probably be Wikilinked.
• Coronal is clearly the more accurate and intended term here, but the usage of dental so entrenched in PIE phonology that even Byrd uses dental over coronal (although maybe he genuinely believes those stops to be dental to the exclusion of alveolar). Sihler surprisingly uses apical, which is neat of course in being ambivalent toward dental/alveolar, but is idiosyncratic among the PIE handbooks. Thinking more about it, maybe it is best to use dental for the table.
• Byrd puts *h₂ and *h₃ in a pharyngeal column and *h₁ in a glottal column. This is neat, but it would be unduly presenting the position of *h₂ and *h₃ as pharyngeal. Ringe places *h₁ in a glottal column as well, but puts *h₂ in the velar column and *h₃ in the labiovelar column. In yet older handbooks you can also find *h₁ being put into the palatal column. It's rather difficult to express without words the scholarly consensus that: *h₂ and *h₃ are in the same place (which could be pharyngeal or uvular), and *h₁ is certainly glottal. I think the more conservative vague presentation is good here, since it'll be much easier to properly deal with the laryngeals with prose.
• Italics – I don't know why the current table in the article italicizes all of the main cells.

Kisseran (talk) 18:18, 12 July 2026 (UTC)

I'm not sold on providing a full table of consonants; I'm not sure what benefit it provides the reader overall. For languages that are living, recently extinct, or have an established written tradition, it makes sense to have one table sit together, but with these distant-past reconstructions where there is substantive debate, it seems at best misleading. The stops should probably be in a 3x5 table (voicing/aspiration x PoA), but for the other categories, it just doesn't really make any sense. Regarding your subnotes:

• Agreed.
• I think dental is the appropriate terminology here both because of WP:COMMONNAME considerations (or whatever the equivalent is in prose) and because there is some literature out there that uses dental/alveolar somewhat interchangeably when the language doesn't distinguish between, for example, /t/ and /t̪/. Similarly, Kapović (2008) calls all stops occlusives (okluzivi) instead of stops (zapornici) or plosives (plosivi). We may note prominent alternative names, but only as footnotes or short parentheticals. Same with the use of the breve or circumflex instead of the acute to mark the palatovelar series.
• Yeah, I don't think it's helpful for us to put the laryngeals in a table at all; I would much rather address them individually in subsections and discuss the differing opinions more or less at length there. I think *h₁ is universally or near universally understood as a glottal (either /h/ or /ʔ/), but *h₂ and *h₃ are much more controversial; I've seen everything from /ɣʷ/ to /ʁ/.
  • As an aside, any thoughts on including *h₄ in the discussion as its own subsection or grouping it with *h₂? Leaning toward giving its own discussion, but I'm sure you have something much more thoughtful to say about it.
• Probably because they're guided by the MOS:LANG guidance since they're not IPA characters per se, but rather orthographic representations. Personally, I think using {{langr}} over {{lang}} is an easy compromise.

ThaesOfereode (talk) 21:04, 12 July 2026 (UTC)

• I think the benefit it provides to the reader is largely the same as the benefit it provides to the reader on most other phonology articles: it's a way to tell what's there at a glance. If you feel comfortable going without a full consonant table, then I would agree that it's best to go without one. Not having consonants as a second-level heading also makes this much easier.
• Thematically, I think it makes the most sense to discuss *h₄ with *a, but I don't see a way of making that organization make sense who doesn't already know enough to not have to read the discussion. In any case, I think *h₄ does not warrant more than a sentence or so of discussion on this article, and any more that isn't explicitly in a historiographical context would be undue weight. Just so we're on the same page, here's my judgment on the reception of some of the more controversial issues being covered:
  • Any number of laryngeals other than three is borderline fringe. I haven't read Zair's works in too much detail, but in my personal experience, he assigns those models more merit than they deserve, and more importantly more merit than usually assigned in the literature. When Mallory & Adams came out it was immediately flamed for endorsing *h₄ in an introductory text, it remains a major flaw of that book.
  • Glottalic theory is an arguably mainstream alternative model of stops. It shouldn't be mentioned here outside of the subsection dedicated to it, but it merits an extended discussion.
  • The two sides to the existence of *b and *a are roughly similar in reception, and should be afforded equal coverage here. Rejection of *b and *a are still strongly associated with Leiden, but proponents of *b and *a tend not to make particularly strong attempts to justify their positions or explicitly remain ambivalent. To be clear, I think your current verbiage with *a is fine.
• I should maybe propose an edit to the lang templates to not italicize text with preposed asterisks.

Kisseran (talk) 21:43, 12 July 2026 (UTC)

• I think we're fine without it, but I won't object if we decide later we need one.
• Controversy replies:
  • I agree regarding h₄'s status as a fringe idea; it's rarely – if at all – discussed in the papers I've read. For what its worth, Zair (2012) gives it one sentence of daylight and resolutely rejects it ("consequently an a-colouring *-h₄- cannot be posited"). Maybe a little more to say at the laryngeal page itself, but this is much more of a primer and it's good to avoid the same pitfalls as M&A. In any event, we'll cross that bridge when we get to it.
  • That sounds right to me; Byrd (2018) writes: "While still maintaining a small group of ardent followers (particularly in Leiden), the [glottalic theory] has lost much of its support today." I think Leiden's overall influence in the field and the number of noteworthy mainstream linguists it produced (Kortlandt and Beekes are chief in my mind) certainly warrants discussion as a "not-fringe-but-not-so-widely-accepted alternative theory".
  • I would say so. If memory serves, *b has been posited as a loan phoneme, but I'm sure that Olander article will provide better insight than my half-present memory of half-read articles. Glad the *a verbiage is good. I was trying to tag anti-*a as non-mainstream, though it was first introduced to me as relatively uncontroversial, interestingly enough.
• I'm not sure we need to go that far; that's what the old {{PIE}} template used to do and it made following MOS:LANG extremely tedious. It's better to just use {{langr}} for the table, which is just a {{lang}} template without italics.

ThaesOfereode (talk) 01:14, 13 July 2026 (UTC)

For some bizarre reason I was under the impression that the typographical convention in general when it comes to presenting reconstructed forms was to set them in roman, not italic. Upon actually checking, it seems that this is clearly not the case, so I don't actually know where I got that idea from (I'm sure there are some works out there with this convention, but it wasn't any of the ones I immediately checked) Kisseran (talk) 05:15, 13 July 2026 (UTC)