The Linguistic Relativity hypothesis e.g., Whorf, 1956, which states that

1
Do Spanish Tables Have Curves? A Semantic Priming
Investigation of Linguistic Relativity Tamar
Degani1,2, Natasha Tokowicz1,2 Brian
MacWhinney2,3 ¹Department of Psychology
Learning Research and Development Center,
University of Pittsburgh ²Center for the Neural
Basis of Cognition, University of Pittsburgh
Carnegie Mellon University ³Departments of
Psychology and Modern Languages, Carnegie Mellon
University
46th Annual Meeting of the Psychonomic
Society November 10-13, 2005 Contact
Tdegani_at_pitt.edu

• The Linguistic Relativity hypothesis (e.g.,
  Whorf, 1956), which states that
• language affects thought, was examined in the
  case of Grammatical
• Gender in Spanish.
• Previous research people who speak a language
  with a grammatical gender system describe,
  categorize, and remember inanimate objects based
  on the objects grammatical gender (for review
  see Boroditsky, Schmidt, Phillips, 2003).
• Does the grammatical gender system guides
  peoples thinking in an on-line task as well?
• Gender priming effects have been shown in the
  past using adjectives or articles as primes and
  nouns as targets - syntactic violation (e.g.,
  Bates, Devescovi, Hernandez, Pizzamiglio,
  1996).
• Can we find evidence of semantic priming for
  words that match in gender?
• We hypothesized that if the grammatical gender of
  a noun becomes part of its semantic
  representation then it would be named faster when
  preceded by a noun that matched in gender
  compared to when preceded by a noun that did not
  match in gender.
• Method Experiment 1
• Task - Primed naming
• Participants
• 24 native Spanish speakers

• Stimuli Experiment 1
• 200 Spanish nouns. Each participant saw 80 pairs
  (counterbalanced).
• Grammatical Gender condition inanimate noun
  prime inanimate noun target (40 pairs).
• Semantic Gender condition animate noun prime
  inanimate noun target (40 pairs).
• Half the pairs matched in gender and half did
  not.
• Stimuli were matched for frequency and length
  across conditions.
• Results Experiment 1
• No effect of matched gender on naming latencies
  was found.
• Matched pairs (91) were named less accurately
  than unmatched pairs (94).

• Preliminary Results Experiment 2
• No effect of matched gender on naming latencies
  was found.
• Adjective-noun pairs (92) were responded to more
  accurately than inanimate-noun pairs (89).
• Discussion
• In an on-line naming task we did not find
  evidence for a semantic effect of grammatical
  gender. The influence of grammatical gender may
  be dependent upon strategy or the participants
  state of mind.
• In contrast to previous research on gender
  priming, we did not find grammatical gender
  priming in the Adjective condition in Experiment
  2. This could be because in Spanish adjectives
  usually follow the noun and not precede it.
  However a post-hoc analysis revealed that
  feminine noun targets (but not masculine targets)
  were processed more quickly when preceded by a
  matched adjective than when preceded by an
  unmatched adjective.
• A post-hoc analysis of Experiment 1 data revealed
  that in the grammatical gender condition
  masculine noun targets were named faster than
  feminine targets, but in the semantic gender
  condition, feminine targets were named faster
  than masculine targets. We are currently
  investigating possible explanations.
• Conclusions and Future Directions