Should I use the historical method?

[mizuz]

I've heard a few "high level" conlangers praise the historical method, that is the method whereby you first create a proto-conlang and then derive your actual conlang by applying the laws of historical linguistics to it. I can see that this can give some depth to the conlang as you have a history of it, but besides that, is there any other merit to this method which is worth the effort?

[do_shahbaz]

It would give 'depth' to derivation and derivational morphology (with the sound-changes and semantical shifts obscuring the latter), and would suit making a language family.

[Dormouse559]

I'll start out by saying you don't need to make a proto-conlang. Conlanging doesn't have rules, so it's a matter of assessing what you personally want to do with your conlang and what techniques would help you accomplish that.

A historically derived, or diachronic, conlang can have some benefits, mainly stemming from that depth you mentioned. For one thing, having a protolanguage makes things much easier if you ever decide to create dialects/sister languages. There's also a one-step removal from the fundamental arbitrariness of conlanging, so you can have ready justifications for the weirdness that's normal in natlangs but hard to emulate in conlangs. Think irregular declensions, which are often the result of regular application of sound changes. If there's a conworld or -culture with much history attached to the conlang, having the previous form(s) of the language can help you give more realistic texture. Related to that, if the conlang borrows from other languages, you can have old borrowings that have been fully naturalized and gone through sound changes alongside recent borrowings that are mostly unchanged from the source (like the doublet petty and petite in English).

Those are just some possible benefits, and many of the features I mentioned could be created without resorting to diachronic methods. Also, I should point out that you don't necessarily have to create a full-fledged, deep-in-the-past protolanguage to get some of these benefits. Even just a general idea of how the language has changed over the past hundred years can be useful.

[mizuz]

Thanks for the answers! So many things I didn't consider.

[linguoboy]

I think the chief benefit is that it will spur you to do research on the process of language change, which will expand your knowledge of what kind of changes are possible (and likely) and which ones result in "naturalistic" results.

We used to play a game where someone would provide a word of a short phrase in a conlang and we'd all take turns applying sound changes to it. Might be fun to revive that again.