What is the correct vowel quantity for the participle of legō?

Question

In the following, vowel quantities which I am uncertain of, will be marked with both a breve and a macron, so they should not be considered the answer; that is what I am searching for.

This whole investigation began with collē̆ctiō. I was surprised to see that both Gaffiot and L&S listed it as collectiō, whereas Wiktionary lists it as collēctiō. The Wikipedia entry lists it as following:

1. collēctiō < collēctus +‎ -tiō, from colligō.
2. colligō < con- +‎ legō
3. legō, legere, lēgī, lēctum

Gaffiot and L&S agree on 1 and 2, but on 3, they list the following:

• Gaffiot: lĕgo lēgi, lectum, ĕre
• L&S: lĕgo, lēgi, lectum

Both agree that it has no clear vowel length; they do not specify it as neither short nor long. I then tried to have a look at the PIE origin of legō via the Wiktionary entry, to see if there could be any clue there, here with the entries relevant for Latin listed:

Root
*leǵ- (imperfective)
to gather, collect, with derivatives meaning to speak

Derived terms
Terms derived from the Proto-Indo-European root *leǵ-

• *leǵ- (root present)
  • Proto-Italic: *legō
    • Latin: legō (see there for further descendants)
• leǵ-s
  • Proto-Italic: *lēks
    • Latin: lēx

This is about where my ability to gather clues and derive any firm conclusions stop. Thus my question: What is the correct vowel quantity for the participle of legō: lē̆ctum? Adding an explanation of why would be most helpful. I am aware of the phenomenon of creating perfect tenses by vowel lengthening, and suspect therefore that the Wiktionary entry in fact is the more precise, but am not well-enough versed in this to arrive at any conclusions myself.

Answer 1

Hidden quantity in long syllables

The convention adopted by Lewis and Short seems to be not to indicate a vowel quantity if the syllable is long anyway. For example, they write făcĭo, fēci, factum, without indicating whether it is făctum or fāctum. In fact, what they indicate with a macron might well be the length of the whole syllable, not just the vowel, and the consonant cluster ct makes it evident that the syllable is long. In cases like this the quantity of the vowel is hidden: no metric considerations will reveal it.

I find this annoying. While knowing the length of each syllable is enough for figuring out whether you are scanning poetry right, it is not all there is to pronunciation. (This might have to do with my native language making a distinction between 'autonsa' and 'autoonsa', so to me hidden quantities make an audible difference in pronunciation.) I would be interested in knowing whether they have deemed hidden quantities just uninteresting or also unknowable. The conclusion is: Don't take the lack of a macron or breve as a sign of anything in L&S if the quantity of the vowel is hidden.

The OLD also gives the past participle as lectum without diacritics. The conventions section at the start explicitly states: "Normally only long vowels in a metrically indeterminate position are marked."

Dictionaries that do have an opinion

Many otherwise good dictionaries don't seem to indicate hidden quantities. But there are dictionaries out there that do, so let's look at them.

• Reijo Pitkäranta's Lexicon Finnico–Latino–Finnicum (or Suomi–latina–suomi-sanakirja) says lēctum (and lēctor and lēctiō).
• The older Finnish dictionary by Adolf V. Streng (Latinalais-suomalainen sanakirja) does the same.
• Wiktionary, as you mention, gives lēctum.

These are all the dictionaries I was able to locate that did state something about the quantity. They all agree, so lēctum it is. I am fully aware that this is not a linguistic proof by any means; this is just an exercise in using dictionaries. Perhaps someone can provide something more etymologically-natured in another answer?

In general

If you need to figure out the hidden length of a vowel, it can be a little tedious to ask here every time. (You are welcome to if you want to!) What you need is to find a reliable dictionary that does indicate these quantities, and use it as reference material. You could even purchase a copy of either of the two dictionaries for Finns I mentioned; you only need to look up the spellings of the Latin words.

If anyone has good suggestions for dictionaries of this kind, especially online, I would be happy to mention them here. Wiktionary is the only one I know to fit the bill, and it is not the most trustworthy of sources.

Answer 2

It's long.

The two consonants after the vowel mean that poetic meter can't tell us anything about the vowel length. However, ē and ĕ had different descendants in Romance: ē became Proto-Romance e, while ĕ became Proto-Romance ɛ.

In stressed syllables, Spanish generally shows e for Romance e, but ie for Romance ɛ: miedo < mĕtum "fear", but deber < dēbēre "owe". And fortunately for us, a derivative of collēctus with appropriate stress survived into Spanish: cosecha < collēcta "gathered (crops)".