What's the difference between 'voiceless' and 'unaspirated'?

Question

It's an additional question to the following:

• Why is a voiced, voiceless unaspirated, and aspirated distinction so rare cross-linguistically?

My understanding

What distinguishes aspirated/unaspirated is

voice onset time, the time between the release of the stop and the start of voicing,

where I understand voicing as the vibration of vocal cords. And unaspirated sounds have close-to-zero VOT and aspirated sounds greater VOT.

From the comments in the linked question, voiced/voiceless distinctions are done on the negative side of VOT. For simplicity (or by normalization), I assume VOT varies between -1 and 1. Then voiced sounds have VOT close to -1; and voiceless ones closer to zero (i.e., VOT greater than voiced ones by usual comparison).

Question

What is the difference of unaspirated and voiceless sounds? As I see in the link, there seem to exist languages that have three way distinctions. In such a language, I suppose the sound with VOT～-1 is called voiced and the sound with VOT～1 is called aspirated. What is the middle one called? Is it like the middle one is called voiceless if the distinction with the voiced sound is made within the negative side¹ and called unaspirated otherwise?

---

¹ I mean, if the language making the 3-way distinction distinguishes sounds of say, VOT(-1, -0.3) and VOT(-0.3, 0.1) and VOT(0.1, 1), where sounds whose VOT is between a and b are meant by VOT(a,b). Then -0.3 is the border which is negative, and VOT(-0.3, 0.1) is called voiceless.

Answer 1

There may or may not be any difference!

As you say, both voicing and aspiration are extremes of voice onset time. Plenty of languages distinguish consonants based on VOT, and most of those languages have a binary contrast. Sounds with a VOT lower than a certain threshold are "lenis" (to avoid using the terms "voiceless" or "unaspirated" for the moment), and with a VOT higher than that threshold are "fortis".

Different languages put this threshold in different places. In Mandarin, it's very high, in Spanish, it's very low, and in English, it's somewhere in the middle. By convention, in Mandarin, we call the fortis consonants "aspirated" and the lenis consonants "unaspirated". In Spanish, we call the fortis consonants "voiceless", and the lenis consonants "voiced". In English, well…it depends who you ask!

In some other languages, like Ancient Greek, there's a three-way distinction in VOT. In these languages, we call the high-VOT one "aspirated", the low-VOT one "voiced", and the one in the middle alternately "voiceless", "unaspirated", "voiceless unaspirated", or just "tenuis".

So, what's the difference between "voiceless" and "unaspirated", acoustics-wise? Not much, necessarily! They both mean "closer to VOT 0". But what they contrast with is different.

(Also, VOT is normally a measure of time, measured in milliseconds or the like. So it's best to report the numbers directly rather than normalizing them.)