Yosef Grodzinsky presented “The analysis of negative quantifiers: multi-modal evidence” as a colloquium speaker at MIT this past Friday.

Recent discussion of  negative  quantifiers  (see  Penka,  2011  and  references  therein) focuses  on  two  main  questions:  do  these  quantifiers  decompose,  and  if  so,  what  are the  mechanisms  for  decomposition?  In  this  talk,  I  will  describe  a  series  of  neuro-­  and psycholinguistic  experiments  my  colleagues  and  I  have  conducted  with  healthy  and   brain-damaged  participants  that  aim  to  provide  relevant  evidence.  These experiments  recorded  responses  from  several  modalities,  as  participants  were analyzing  sentences  with  positive  and  negative  proportional  and  degree  quantifiers   in  German  and  English  (e.g.,  mehr/weniger-­als-­die-­Hälfte  die  Kreise  sind  Gelb, many/few  of  the  circles  are  blue).

All  experiments  used  a  Parametric  Proportion  Paradigm  (PPP):  participants were  exposed  to  sentence-scenario  pairs,  and  were  requested  to  make  truth  value judgments.  Sentences  contained  a  quantifier  in  subject  position,  and  scenarios   depicted  a  proportion  between  2  types  of  objects.  Proportion  was  a  parameter, systematically  varied  across  images  that  were  presented  with  each  sentence  type.

Our  first  experiment  used  functional  MR  imaging  to  extract  a  signal  that represents  localized  brain  activity.  It  aimed  to  identify  brain  loci  that  evince  an intensity  differential  between  the  contrasting  stimuli.  Signal  intensity  for  sentences with  negative  quantifiers  was  higher  than  that  for  their  positive  counterparts  only  in Broca’s  region.  Importantly,  no  other  localizable  intensity  contrasts  were  found.

A  second  experiment  (currently  only  a  pilot)  confronted  English  speaking, focally  brain  damaged,  Broca’s  aphasic  patients  with  the  same  task.  However  here, the  dependent  measure  was  error  rate.  The  results  suggest  a  remarkably  selective   deficit:  While  patients  performed  near-normally  on  the  positive  quantifiers,  their scores  were  drastically  reduced  when  the  stimuli  contained  negative  quantifiers.

A  third  experiment  attempted  to  take  a  deeper  look  at  the  behavioral signature  of  quantifier  analysis  through  a  study  of  complex  RT  functions  obtained from  healthy  participants.  Here,  too,  we  observed  that  the  signature  of  negative quantifiers  is  quite  distinct  from  that  of  their  positive  counterparts.

In  this  talk,  I  will  try  to  connect  these  results,  obtained  through  different modalities  from  different  populations,  to  previous  ones  that  come  from  parametric studies  of  overt  syntactic  movement  with  healthy  participants  in  fMRI,  and  with Broca’s  aphasic  patients.  I  will  propose  that  a  generalization  over  the  experimental results  supports  an  analysis  of  sentences  with  negative  quantifiers  that  assumes covert  movement.  I  will  then  try  to  situate  these  results  in  the  broader  context  of  a research  agenda  that  tries  to  create  a  brain  map  of  syntactic  and  semantic knowledge.