Ease of L2 acquisition of SOV and SVO/VSO word order

Question

Is it easier for speakers of SOV languages to learn SOV/VSO syntax or vice versa?

I've been reading about constructed languages recently, specifically the ones intended to be easy to learn such as Esperanto and Ido or trivial to learn but not very expressive such as Toki Pona. These languages or vaguely-language-like-things are basically head initial and have SVO as their dominant word order. I'm curious as to whether this is the "best" choice if the goal is to be as friendly as possible to L2 learners.

OV and VO languages are approximately equally common, although things might be different if we counted languages by number of speakers.

So, I'm curious which order is easier to learn for native speakers of languages using the other order... or if perhaps word order is totally insignificant when it comes to L2 acquisition and ease or difficulty is dominated by other factors.

So far, I've seen a couple of works about L2 acquisition of specific features in specific languages by English speakers and some theoretical things such as this paper that seek to explain the presence of case-marking in SOV languages.

Answer 1

Many SOV languages are head final. Korean and Japanese are famous examples. In head final languages, sentences that begin with a subordinate clause are temporarily ambiguous. In other words, the listener or reader cannot be certain if the words encountered belong to the main clause or the subordinate clause.

For example,

"The book that we bought yesterday in town fell on the floor."

At no point in the sentence is there any ambiguity.

The same sentence in a head final SOV language would be,

"We town in yesterday bought that the book the floor on fell."

Here there is initial ambiguity about whether constituents such as "we", "town in", and "yesterday" belong to the main clause or to the relative clause. In fact, even when one gets to "the book" there is nothing that assures the decoder that he or she is now in the main clause. One could be at the beginning of another embedded relative clause. The ambiguity is not resolved until the main verb at the end of the sentence is reached.

Intuitively it can be assumed that this structural ambiguity causes significant cognitive load. Eye tracking experiments have shown that indeed this is the case, suggesting that SOV languages are harder for SVO language speakers to learn than SVO languages are for speakers of an SOV language.

See "COGNITIVE AND LINGUISTIC FACTORS AFFECTING SUBJECT/OBJECT ASYMMETRY: AN EYE-TRACKING STUDY OF PRENOMINAL RELATIVE CLAUSES IN KOREAN" Language Vol. 86, No. 3 (SEPTEMBER 2010), pp. 546-582 (37 pages) Published by: Linguistic Society of America.

Answer 2

From what I can gather based on the article cited below, left-branching structures are more accessible than right-branching ones to speakers with the opposite branching direction. This is the opposite result of what I expected; I thought right-branching languages would just be easier to parse across the board, but not by a significant amount.

The article points out that users of right-branching languages can make parsing decisions immediately and users of left-branching languages must delay these decisions -- or make speculative decisions using statistical or semantic information.

If there is indeed a difference between the strategies used by left-branching and right-branching language users for parsing, or a difference in the extent to which strategies from a common pool are used, then it is possible that the strategies favored by left-branching speakers will not carry over as well to a novel language. Or they might carry over better.

This article, published in Nature, is about an experiment exploring the differences in working memory performance for initial and final stimuli in speakers of left and right-branching languages. It contains the following remark as part of its background material before the experiment.

Speakers seem to develop a "bias" toward the branching direction more common in their language, so that LB structures are harder to process for RB speakers (due to the higher working memory needed to retain the intermediate products of computation 67,68,69; see 70 for experimental evidence), but they are more accessible than RB structures for LB speakers 65, 71, 72.

If I'm interpreting the above sentence correctly, this means that speakers of right-branching languages are better able to interpret left-branching structures than vice versa.

Here are the references cited in that sentence (and 66 because bulleted lists are forced to be numbered consecutively).

65. Mazuka, R. & Lust, B. In Proceedings of NELS 18 (eds Blevins, J., Cart, J.), 333–356 (Univ. of Massachusetts, Amherst, 1988).

66. Pienemann, M. (ed.) Cross-linguistic aspects of Processability Theory (John Benjamins Publishing CO, Amsterdam, 2005).

67. Frazier, L. & Fodor, J. A. The sausage machine: a new two-stage parsing model. Cognition 6, 291–325 (1978).

68. Gibson, E. Linguistic complexity: locality of syntactic dependencies. Cognition 68, 1–76 (1998).

69. Kemper, S. & Kliegl, R. (eds) Constraints on language: aging, grammar, and memory (Kluwer Academic Publishers, Boston, 2002).

70. Friederici, A. D., Chomsky, N., Berwick, R. C., Moro, A. & Bolhuis, J. J. Language, mind and brain. Nat. Hum. Behav (2017).

71. Lust, B. & Mazuka, R. Cross-linguistic studies of directionality in first language acquisition: response to O’Grady, Suzuki-Wei and Cho, 1986. J. Child Lang. 16, 665–684 (1989).

72. Lust, B. (ed.) Studies in the acquisition of anaphora (Kluwer, Boston, 1986).