What, if any, aspects of linguistic typology affect the percentage of loanwords that a language has in its lexicon?

Question

By "loanword," I'm referring to words borrowed into language X from other languages and altered only as the phonology of language X requires. Examples would include "le weekend" in French, "das Internet" in German.

What, if any, aspects of linguistic typology affect the percentage of loanwords that a given language has in its lexicon?

Obviously, many other factors affect the percentage of loanwords in a language's lexicon. I can think of a few off the top of my head. How much contact do the speakers of language X have with the speakers of other languages? Are the speakers of language X a minority culture in the area where language X is spoken? How often to the speakers of a language calque foreign terms instead of just adapting foreign words to language X phonology?

My question is, are there any instances in which language typology influences the percentage of loanwords in a given language? For example, all other things being equal, would a language with a lexicon dominated by bound morphemes have fewer loanwords than a language with a lexicon dominated by free morphemes?

Answer 1

I don't think whether a language has many bound morphemes or not (or whether it is synthetic or analytic) has much influence on how readily loanwords are accepted into the language. Empirically showing that this is true would be complicated due to the many other factors you mentioned that influence the number of loanwords.

To provide some anecdotal evidence I'll give examples of recent or ad-hoc borrowings from two synthetic languages and how they are integrated into the complex morphology of these languages. These examples show that languages with many bound morphemes readily borrow words and add native morphemes to them.

Polish

I recently heard the word niezasejwowany in a Polish pop song. This is a past participle and ultimately borrowed from English to save. It appears to be so recent a loanword that it's not in the rather good dictionary I use.

Analysing niezasejwowany gives nie-za-sejwowa-ny, un-PERF-save-PP-Sg.m.

• -ować is a very productive verbal suffix usually used with borrowed verbs -> sejwować
• za- is a perfective marker -> zasejwować
• past participle singular male is zasejwowany
• and adding the negative prefix nie- gives niezasejwowany

Although this appears to be a rather recent loanword it is readily integrated into Polish morphology.

German

A while ago I heard a shop assistant praise another's hard work:

Du bist aber heute wieder hart am Worken

You are but today again hard at working/work

You are really working hard today

Worken is a deverbal noun from what I assume must be the verb worken, obviously borrowed from English to work. I haven't heard anyone use the verb worken in German, so I assume this is an ad-hoc borrowing or code-switching. The crucial point is that even such an ad-hoc borrowing is readily integrated into German morphology and gets a deverbal -en suffix to derive a noun.

Conclusion

Based on these examples I would say that even very recent borrowings are integrated into the morphology of synthetic languages without any problems. And if this is the case the number of bound morphemes a language uses should have no influence on how many loanwords it accepts.