Benjamin Bruening (2018), Language (Linguistic Society of America) 94(1), 1–42 — DOI-10.1353/lan.2018.0000
Key takeaways
- The lexicalist hypothesis, that a separate, encapsulated component builds words before phrasal syntax, is wrong and superfluous. Phrasal syntax feeds word formation, phrasal operations reach inside words, and the two obey the same principles. The facts it was meant to explain have independent explanations that never mention “word”, so grammar needs only one combinatorial component.
Seeded from the abstract and a skim, before the reading in NOTE-065. What follows is what the work says about itself.
The lexicalist hypothesis holds that the grammar component producing words is distinct and strictly separate from the one producing phrases. Bruening argues it is wrong for three reasons: phrasal syntax often feeds word formation, phrasal syntax can access subword parts, and claims that word formation and phrasal syntax follow different principles are false. It is also superfluous, because the facts it was supposed to explain have independent explanations. The most parsimonious model follows: one combinatorial component that builds both words and phrases.
Standing in the record
Filed from the survey of 2026-09-25 of work the anthology set aside as out of scope (tier C): 380 seconds of active reading over 2 sessions in the papers-feed tracker. It was filed Deferred, unread. NOTE-065 is the close reading of 2026-09-25, and it placed the work: Active — This is the most complete single-article case against lexical integrity from inside generative syntax, and it is the syntactic source LIT-031 cites; read it together with Müller’s 2018 reply in the same volume, which disputes its extraction, coordination and resultative arguments.
When seeded, before the reading in NOTE-065, the entry gave a priority for a deeper reading: medium — 380 s over 2 sessions. The argument is fully laid out in the skim, and the value of a deeper read is in the published replies rather than in re-reading this article.
What the seeding asked a deeper reading to check (NOTE-065 is that reading):
- This is the theoretical-syntax side of c61’s typological argument and the main linguistic source for c62. Together they argue there is no privileged word level, which is relevant to how subword tokenization and compositional generalization in language models are discussed.
- The same Language issue apparently carried responses (a ResearchGate hit shows Müller’s “The end of lexicalism as we know it?”). A deeper read should find and weigh the replies.
- Check §5’s alternative explanations, where the independent-principle accounts are least worked out; the author defers other languages to future research (footnote 28).
Access when seeded: Crossref (print issue March 2018, no day, so the published date uses YYYY-MM-01; Crossref also carries a spurious 2026-01-01 “online” date). Project MUSE returned a verification page. I read the published version as a Project MUSE PDF mirrored on a UCSC course page (babel.ucsc.edu, 43 pp.; the author’s own copy is at udel.edu/~bruening/Downloads/LexicalismSuperfluous4.pdf): abstract, keywords, all section heads, §2.1, §3.1, §5 opening, §5.6 and §6 conclusion. I did not check whether an earlier manuscript circulated, for example on LingBuzz.