Question about PIE and I guess language complexity in general

[Xhin]

A bunch of related but different words across multiple language families come from single PIE roots. Was PIE a simpler context-based language, or were its other useful words just lost over time?

Also, I know that PIE is a backwards construction rather than a clearly defined language. I guess my question could apply to any language not that far back in the etymology tree that had a less complex civilization behind it.

My intuitive thinking here is just that the complexity of a language increases as the civilization around it does. A word like "join" has all kinds of contextual meanings, and if the language wasn't standardized, I could see how these would shift into multiple different words over time.

That said, I can't imagine any language working where, for example, "deru" means "tree", "tray", "trust", "truth", "durable", "shelter", "during", "trough" and "endurance". There are some roots that are even worse.

[Ketsuban]

Almost certainly there were words spoken by the most recent common ancestor of the Indo-European language family which have not survived to the present day.

That said, I can't imagine any language working where, for example, "deru" means "tree", "tray", "trust", "truth", "durable", "shelter", "during", "trough" and "endurance".

This is a misunderstanding. The root *drew- or *deru- doesn't have all those meanings in the ancestral language—different communities speaking descendants of the ancestral language either inherited or innovated words like *dóru, *dérwis, *derwóm, *drewóm, *dréwnom, *derdrewom, *drukós etc. produced through their own derivation processes, which then diverged in meaning over time. (You've included a bunch of derivations from Latin durus, but it's not clear if that's derived from this root or not.)

[Travis B.]

Since when are languages from more 'developed' societies more complex? There is nothing that substantiates this idea.

While languages from more 'developed' societies may have more specialized technical vocabulary than less 'developed' societies, the opposite is often true about vocabulary about the natural world -- less 'developed' societies often have far more vocabulary in common use about the natural world than more 'developed' societies.

After all, if you actively gather plants from the wild you need to know what is good to eat or is otherwise useful and what is deadly poisonous, a need which is less acute for those of us who primarily get things from the grocery store.

[zompist]

A bunch of related but different words across multiple language families come from single PIE roots. Was PIE a simpler context-based language, or were its other useful words just lost over time?

I think what you're noticing is not complexity, but the process of word creation. Almost always, people create words from other words (possibly foreign ones of course). Probably the commonest technique is metaphor. When doing etymological research for the Conlanger's Lexipedia, I found that the standout is stand, which (in one language or another) underlies English constant, assist, distant, exist, insist, instant, rest, stable, stage, stanza, state, station, statue, stay, steady, understand, as well as non-English words meaning always, admit, be, big, brave, build, can, cease, finger, form, happen, help, hour, immediate, last, old, place, rice, room, ship, soldier, tail, thick.

Why? Well, as a conlanger it should be a obvious: stand is a really useful idea, which you can build into words for staying, constancy, taking up space, being erect. There may be a chain of metaphors; e.g. take up space > take up time > being old.

All this doesn't mean that earlier languages had fewer words. Just look at how many words were replaced in Romance.

My intuitive thinking here is just that the complexity of a language increases as the civilization around it does.

Well, yes and no. You get more technical terms; written language also preserves old words that would otherwise be lost.

But no, "primitive" people don't have "primitive" languages. Learn a few and you'll see.

For balance I should note the other intuitive view people often have, which is that languages get simpler over time— e.g. Latin is more morphologically complex than French; Old English than English. Though both view are wrong, they may be truer taken together: there are processes at work in both directions, toward simplicity and toward complexity, and they tend to balance out.

[Travis B.]

All this doesn't mean that earlier languages had fewer words. Just look at how many words were replaced in Romance.

This is well-illustrated by contrasting what we know of Classical Latin with purely-reconstructed Proto-Romance that makes no reference to written Latin; they differ quite a bit.

[Darren]

We can only reconstruct into PIE words which appear in numerous branches. Statistically, this mostly means basic concepts. More specific terms, which undoubtedly existed, have been subject to more replacement in descendants, either by language-internal formations (usually from said basic concepts) or by borrowing. When enough major branches do this, we can't reconstruct the word to PIE any more. Think how many words there are in any branch of IE with uncertain etymology, that could well derive from a specific PIE root which was simply lost in most other branches.

PIE has a massive known vocabulary compared to most other proto-languages. For TNG (at least some of which is legit), we know just about the words for "louse", "egg" and a couple of pronouns, but that doesn't mean they just said "me louse you egg, heh heh?" all the time.