False friends: the effects of mismatching in morphological complexity

False friends: the effects of mismatching in morphological complexity on a backward lexical
translation task
VICTORIA JANKE and MARINA KOLOKONTE
University of Kent
This study contributes to the question of the degree to which learners of a second language rely on
L1 when processing their L2. It does this by focusing on false cognates, words which share similar
form, but have different meanings. These words trigger translation errors in learners who, for
example, confuse the French ‘librairie’ with English ‘library’ instead of accessing the correct
translation ‘bookshop’. Using a backward lexical translation paradigm (L2 → L1), we manipulate the
levels of morphological complexity of false-cognate pairs and examine the repercussions this has on
the false-friend effect. False-cognate pairs were created from French (L2) and English (L1). 60
participants with differing levels of proficiency (A Level; Degree Level; Adult Learners; Bilinguals)
were tested on mono-morphemic items (simplex), poly-morphemic items (complex), items whose
morphological complexity in French exceeded that of their English counterpart (mismatch), and
control items. Our expectation was that the more levels at which a word in L2 and L1 match in form,
the more a speaker would be misled and thereby translation success rate would decrease.
Translation success rate followed a uniform pattern: control > mismatch > simplex > complex, for all
groups regardless of proficiency. A Friedman test confirmed the different scores between these
conditions to be significant (p<0.001), and the Pages’s L Trend test, conducted to confirm the
predicted order of accuracy (control > mismatch > simplex > complex) (p<0.001). With respect to the
false-friend effect, the pattern of responses for all groups was also uniform: complex > simplex >
mismatch. A Friedman test gave significance (p<0.001), and a follow-up Page’s L Trend test to
confirm the predicted order (complex > simplex > mismatch), was also significant (p<0.001). Our
results show that morphological complexity does impede translation, and we discuss these effects,
and their implications for previous studies in this area, in particular Gordon (1989), Carroll (1992)
and Smith and Tsimpli (1995). A pilot test on two profoundly deaf learners of French performing the
same test are also included, with a view to indicating how morphological effects might be further
separated from phonological effects. Progress in our understanding of the false-friend effect is
important, as it sheds light on the way in which words are stored and accessed in the bilingual’s
lexicon (Dijkstra et al 2000, Lemhöfer and Dijkstra 2004, Longtin and Meunier 2005, Marslen-Wilson
1994).