How Simple Complex Words Can Be: Morphological Processing and Word Representations

Abstract

In four lexical decision experiments we studied the effect of morphological complexity on word recognition. Some potentially relevant linguistic aspects of derived nouns were varied: the location of the affix (prefix vs. suffix); the genuineness of the affix (real vs. pseudo); the orthographic legality of pseudo-stems; semantic compositionality; the nature of the stem (free vs. bound); the origin of the complex word (Latinate vs. Germanic); the currency of the stem (current vs. moribund). Furthermore, in the first two experiments, we systematically varied the proportion of complex and simple words to see whether strategies influence morphological effects on recognition times. Consistent with Taft's notion of affix stripping, pseudoprefixed words show longer mean decision times and higher error rates than truly prefixed words. Further, the prefixed and nonprefixed (but not pseudoprefixed) words are processed equally rapidly, indicating that a decompositional process is efficient. No differences were found, however, for suffixed, pseudosuffixed, and nonsuffixed words. There was no effect of the proportion of simple and complex words. There are some indications that the etymological origin of words may affect recognition times, but no other linguistic aspects of derivations do so. The results of the four experiments are interpreted as supporting a left-to-right process for word recognition in which morphemes are extracted automatically. During this process information encoded by morphological structure becomes available for other processes.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 60 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.