That thou art blamed shall not be thy defect, For slander’s mark was ever yet the fair; The ornament of beauty is suspect, A crow that flies in heaven’s sweetest air. So thou be good, slander doth but approve Thy worth the greater, being wooed of time; For canker vice the sweetest buds doth love, And thou present’st a pure unstained prime. Thou has passed by the ambush of young days, Either not assailed, or victor being charged; Yet this thy praise cannot be so thy praise. To tie up envy evermore enlarged. If some suspect of ill masked not thy show, Then thou alone kingdoms of hearts shouldst owe.

That thou art blamed shall not be thy defect, For slander’s mark was ever yet the fair; The ornament of beauty is suspect, A crow that flies in heaven’s sweetest air. So thou be good, slander doth but approve Thy worth the greater, being wooed of time; For canker vice the sweetest buds doth love, And thou present’st a pure unstained prime. Thou has passed by the ambush of young days, Either not assailed, or victor being charged; Yet this thy praise cannot be so thy praise. To tie up envy evermore enlarged. If some suspect of ill masked not thy show, Then thou alone kingdoms of hearts shouldst owe.

That thou art blamed shall not be thy defect, For slander’s mark was ever yet the fair; The ornament of beauty is suspect, A crow that flies in heaven’s sweetest air. So thou be good, slander doth but approve Thy worth the greater, being wooed of time; For canker vice the sweetest buds doth love, And thou present’st a pure unstained prime. Thou has passed by the ambush of young days, Either not assailed, or victor being charged; Yet this thy praise cannot be so thy praise. To tie up envy evermore enlarged. If some suspect of ill masked not thy show, Then thou alone kingdoms of hearts shouldst owe. Beautiful Light-colored Just Item used as decoration Quality making something beautiful Suspicion Worthy of suspicion Thing suspected Prove Condone First thing Prime of life Springtime Made bigger Set at liberty Uniquely Without sharing Own Carry a debt

“Scholarly glosses, particularly those for the sonnets, have commonly done a disservice both to readers and poems by ignoring the obvious fact that verse exists in time, that one reads one word and then another. A word or phrase can be incomprehensible at the moment it is read and the be effectively glossed by the lines that follow it; a word or phrase can (and in the sonnets regularly does) have one meaning as a reader comes on it, another as its sentence concludes, and a third when considered from the vantage point of a summary statement in the couplet.” -- Stephen Booth, Shakespeare’s Sonnets So when do these ambiguous meanings come in? Booth often glosses them as simultaneous – the word literally means two things at once. Is this consistent with what goes on in the brain when we approach a sonnet?

The “ambiguity advantage”: In visual lexical decision tasks, ambiguous words have faster reaction times than unambiguous words. >

Lexical ambiguity Homonymy (multiple unrelated meanings) Polysemy (multiple related senses) Polysemy confers the ambiguity advantage, but homonymy actually slows RT (Rodd et al., 2002)

Single-entry vs. separate-entry models of polysemy DOOR 1) Aperture in a wall 2) Piece of wood or other material that covers said aperture DOOR (1) Aperture in a wall DOOR (2) Piece of material covering a wall aperture

Questioning Rodd et al. The evidence supports a single-entry model Because this was a behavioral study, the results can only speak to later-stage (postlexical) processing MEG studies can be used to determine the significance of the Rodd et al. results – do they show a difference between homonymy and polysemy processing that exists only in postlexical processing (consistent with the separate-entry model), or one that exists throughout?

M350 MEG response component peaking at ms Reliably responds to lexical stimuli, whether visually or auditorially presented Embick et al. (2001): M350 is earlier for frequent than infrequent words, and M350 frequency advantage correlates with RT frequency advantage, though M350 occurs hundreds of milliseconds earlier -- thus it’s impossible to tell whether M350 is related to lexical or postlexical processing

Pylkkänen, Stringfellow, and Marantz (2002) “Neuromagnetic evidence for the timing of lexical activation: an MEG component sensitive to phonotactic probability but not to neighborhood density” Goal: test whether M350 represents lexical or postlexical processing “If stimulus property A has a facilitatory effect on lexical access and stimulus property B an inhibitory effect on postlexical processing, we can determine whether a given response component reflects pre- or postaccess processing by testing whether it shows facilitation or inhibition for a stimulus that has both A and B.” For this study, stimulus property A is high phonotactic probability (high frequency in the language of the sounds and sound sequences in the word) and B is high neighborhood density (high number of words in the language that sound similar to the stimulus word)

Study Design Block design MEG study Four categories of 70 stimuli each: 1.high probability/density words (BELL, LINE) 2.low probability/density words (PAGE, DISH) 3.high probability/density nonwords (MIDE, PAKE) 4.low probability/density nonwords (JIZE, YUSH) Two blocks of 140 stimuli each, with a pause in between

Results High probability/density stimuli had longer RTs and shorter M350 latencies than low probability/density stimuli, for both words and nonwords “Our results indicate that M350 latencies vary independently from reaction times when stimuli are simultaneously varied along a dimension that affects lexical activation and a dimension that affects selection/decision. This result can only be explained by lexical accounts of the M350; if the M350 reflected postlexical processing, its latency should reflect inhibitory effects of neighborhood density rather than earlier facilitatory effects of phonotactic probability…”

Beretta, Fiorentino, and Poeppel (2004) “The effects of homonymy and polysemy on lexical access: an MEG study” Goal: Use M350 to test single-entry vs. separate-entry models of polysemy Since homonymy is known to be separate-entry, if polysemy is also separate-entry, there should be an RT difference between homonymous and polysemous words (per Rodd et al.) – i.e. a postlexical difference – but no M350 difference – i.e. no difference at earlier stages of lexical processing Single-entry polysemy should show differences between homonymous and polysemous words at all stages of processing (M350 and RT should both be different)

Study Design Block design MEG study Four categories of 32 stimuli each: 1.Single meaning, few senses (“ant”) 2.Single meaning, many senses (“mask”) 3.More than one meaning, few senses (“calf”) 4.More than one meaning, many senses (“bark”) 64-item practice session, four blocks of 42 stimuli (10 initial for practice), break in between blocks Meanings Senses

Results Words with multiple meanings had longer RTs, words with more senses had shorter RTs Words with multiple meanings had a later M350 peak, words with more senses had an earlier M350 peak (there was also a marginal effect of polysemy on M350 peak amplitude) Processing of homonymous and polysemous words is thus distinct for both earlier and later lexical processing

“Polysemy and homonymy yield distinct processing profiles not only in behavioral responses occurring around 600–650 ms, but also in neural M350 responses occurring approximately 300 ms earlier. These results support a single-entry account of polysemy and, conversely, provide no support for a separate-entry account, that is, for an account of lexical ambiguity which claims that both homonymy and polysemy involve multiple lexical entries at some stage of processing. ” So words with multiple meanings have separate, competing entries, but multiple senses have single, simultaneously- accessed entries.

Questions Where is the disjunction between lexical and postlexical processing? Where is the disjunction between homonymous and polysemous words? How does this relate to the idea of a “word” as a collection of features? What does this mean for our experience of reading literature?

That thou art blamed shall not be thy defect, For slander’s mark was ever yet the fair; The ornament of beauty is suspect, A crow that flies in heaven’s sweetest air. So thou be good, slander doth but approve Thy worth the greater, being wooed of time; For canker vice the sweetest buds doth love, And thou present’st a pure unstained prime. Thou has passed by the ambush of young days, Either not assailed, or victor being charged; Yet this thy praise cannot be so thy praise. To tie up envy evermore enlarged. If some suspect of ill masked not thy show, Then thou alone kingdoms of hearts shouldst owe. Beautiful Light-colored Just Item used as decoration Quality making something beautiful Suspicion Worthy of suspicion Thing suspected Prove Condone First thing Prime of life Springtime Made bigger Set at liberty Uniquely Without sharing Own Carry a debt