Left-Corner Parsing and Psychological Plausibility

It is well known that even extremely limited centerembedding causes people to have difficulty in comprehension, but that left- and right-branching constructions produce no such effect. If the difficulty in comprehension is taken to be a result of processing load, as is widely assumed, then measuring the processing load induced by a parsing strategy on these constructions may help determine its plausibility as a psychological model. On this basis, it has been argued [AJ91, JL83] that by identifying processing load with space utilization, we can rule out both top-down and bottom-up parsing as viable candidates for the human sentence processing mechanism, and that left-corner parsing represents a plausible alternative.Examining their arguments in detail, we find difficulties with each presentation. In this paper we revise the argument and validate its central claim. In so doing, we discover that the key distinction between the parsing methods is not the form of prediction (top-down vs. bottom-up vs. left-corner), but rather the ability to instantiate the operation of composition.