Priming

1
Priming

• Eva M. Fernández
• Queens College Graduate CenterCity University
  of New York

2
Priming in lexical access

• Sequence of related words results in faster
  response times (e.g., in lexical decision tasks)
  than sequence of unrelated words
• DOCTOR (as target) will be primed by (non)words
  like
• nurse, hospital
• toctor (non-word), proctor (maybe?)
• ORANGE will be primed by oradge, but PIN will
  not be primed by pit (neighborhood density)
• Lexicon is organized by frequency, as well as by
  form (lemmas) and meaning (lexemes)

3
Tips-of-the-tongue (TOTs)

• Gollan Brown (2006)
• Laboratory induced TOTs with easy and hard
  words
• Variables examined age, bilingualism
• Expt. 1 increased age leads to more TOTs for
  difficult but not for easy targets
• Expt. 2 bilinguals experience more TOTs than
  monolinguals for easy targets, fewer TOTs for
  difficult targets
• Data explained via a two-step process for lexical
  retrieval

4
Structural priming

• Intriguing application Using corpus analysis to
  identify the idiolect of authors and thereby
  confirm (or disconfirm!) authorship
• Requisite assumption priming of some sort?
• Al hablar, al escribir, o mientras pensamos, no
  sólo elegimos las palabras las palabras también
  nos buscan. O, por mejor decir, es el cerebro
  quien las busca tras haberlas guardado de un modo
  particular (Madrigal, 2004, p. 219,
  http//www.h-net.org/cervantes/csa/artics04/madri
  gal.pdf)
• Also linked (possibly) to studies of alignment
  from discourse analysis/sociolinguistics
  perspective

5
Structural priming

• An interesting way to examine the question of
  language independent (unitary) or language
  dependent (dual) sentence planning mechanisms
• How does priming occur?
• Residual activation of prime affects production
  of target
• What time do you? v. At what time do you?
  (Levelt Kelter, 1982)

6
Explaining structural priming

• Structures (in English) that have been shown to
  cause priming
• Dative alternation
• Mary gave a book to her boyfriend / Mary gave
  her boyfriend a book
• Mary donated her inheritance to an orphanage /
  Mary donated an orphanage her inheritance
• Passivization
• Mary read the book / The book was read by Mary

7
Explaining structural priming

• Lexically driven, or beyond the lexicon?
• by-locatives prime by-passives
• The 747 was landing by the airport tower.The
  banana was eaten by the gorilla.
• to-locatives prime to-datives
• The widow drove a BMW to the church.The gorilla
  gave the banana to Mowgli.
• (Bock Loebell, 1990)

8
Explaining structural priming

• Perhaps treelets in a phrase-structure grammar
  are whats activated
• E.g., VP ? V NP PP v. VP ? V NP NP
• Bock (1989) do you get dative priming with the
  following?
• The cheerleader saved a seat for her
  friend YESThe cheerleader gave a seat to her
  friend YES
• Susan brought a book to Stella YESSusan
  brought a book to study NO (difft treelet)

9
Explaining structural priming

• In a lexical entry
• Lemma form representations
• Lexeme essential meaning representation
• Lemmas (for verbs, presumably!) contain
  combinatorial nodes representing what syntactic
  categories can be combined with specific verbs
  (Hartsuiker et al., 2004 Pickering Branigan,
  1998), e.g., NP_NP NP_VP
• So is it so structural? Its all activation of
  nodes in the lexicon!
• Lexical boost effect (Pickering Branigan,
  1998) priming is much stronger when verbs are
  identical in prime and target (e.g., give/give v.
  give/hand)
• Translation equivalence boost (Schoonbaert et
  al.) ditto, for translation equivalent verbs
  (e.g., give/geven v. give/aanreiken)

10
Priming passives in SP/EN bilinguals

• Hartsuiker et al. (2004)
• 24 Spanish native speakers, English as L2
• 22 months in UK, on average (2mos 7 yrs)
• 15 used English at home more than Spanish
• 1 bilingual confederate
• 32 cards depicting an entity performing an action
  and another undergoing action, with a verb
  printed at the bottom
• Animacy of patients 16 animate, 16 inanimate
• Agents always placed on right (location matters
  Marcus, lets talk about this!)

11
Priming passives in SP/EN bilinguals

• 32 targets 96 fillers (actions describable with
  structures other than target) 128 cards for
  naïve participant
• Participant picks up a card, produces a sentence
  (in English), thinks that the confederate is
  checking for meaning
• 128 cards for participant to check against
  confederates speech
• Confederate reads scripted text (in Spanish),
  participant checks for correspondence to picture
  (distractor task)
• Primes are (sic)
• Actives El taxi persigue el camión
• Passives El camión es perseguido por el taxi
• Intransitives El taxi accelera
• OVSs El camión lo persigue un taxi

12
Priming passives in SP-EN bilinguals

• Actives El taxi persigue el camión
• Passives El camión es perseguido por el taxi
• Intransitives El taxi accelera
• OVSs El camión lo persigue un taxi

13
Priming, Dutch/English bilinguals

• Desmet Declerq (2006, Journal of Memory and
  Language, 54)
• 30 Dutch native speakers 30 Dutch/English
  bilinguals(Drawn from same population?
  Difference between groups?)
• 24 targets, 24 primes, 51 fillers, in a written
  sentence-completion task
• For native speakers, everything in Dutch
• For bilinguals, about half of the materials in
  English targets in English, primes in Dutch

14
Priming, Dutch/English bilinguals

• Targets (half plural N1, number agreement
  disambiguates)
• John ontmoette de bazin van de beniendes
  die..
• John met the boss of the employees who
• Primes
• Die politie ondervroeg de veroorzaakster van het
  ongeval die.
• Die politie ondervroeg de veroorzaakster van het
  ongeval dat.
• Die politie ondervroeg de veroorzaakster van het
  ongeval nadat.
• (The police interrogated the causer of the
  accident that/after)
• NB The relative pronoun die is used with
  masculine, feminine and plural antecedents (as
  the determiner de). The relative pronoun dat is
  only used with neuter antecedents (as the
  determiner het).

15
Priming, Dutch/English bilinguals
16
Priming, L2 ? L1 productive

• Conclusions
• Shared syntax? (representations v. processing
  mechanisms)
• All lexically driven?
• More empirical evidence needed!