lisi.beyersmann
A number of studies have shown that skilled
readers decompose morphologically complex
words upon encountering them
(for a review, see
Rastle & Davis, 2008)
. It has been proposed that
this segmentation process is early and automatic
and is driven by orthographic form, while being
blind to semantic content, thus also called
morpho-orthographic
(Rastle, Davis, & New, 2004;
Taft, 2003)
. One key finding in favor of this
proposition comes from masked morphological
priming: the recognition of a target word is
facilitated when it is preceded by a
morphologically related word prime (teacher-TEACH).
Facilitation has also been found in a number of
languages for targets preceded by pseudocomplex
word primes that is words that appear to have a
morphologically complex structure, but are
simplex words (corner-CORN). Moreover,
facilitation has as well been observed from complex
pseudoword primes, that is a non-existing
combination of a stem and affix (flexify-FLEX). For
non-morphological nonword primes, that is a
non-existing combination of a word and a
nonmorphemic ending (flexint-FLEX), mixed results
have been obtained
(Longtin & Meunier, 2005;
Morris, Porter, Grainger & Holcomb, 2011)
.
Recent evidence from French points to a
moderating role of language proficiency: the magnitude
to which morpho-orthographic information is
used increases as a function of individual
vocabulary and spelling skills in adults
(Andrews &
Lo, 2013; Beyersmann, Casalis, Ziegler &
Grainger, 2014)
.
Only few studies have been concerned with
morphological decomposition in beginning
readers. The few studies from English and French
used complex word primes, pseudocomplex
word primes and non-morphological word
primes.
Quémart, Casalis and Colé (2011
) found
priming in French grade 3, 5, and 7 children
from complex as well as pseudocomplex words,
thus suggesting that children already use
adultlike decomposition processes. In contrast,
Beyersmann, Castles and Coltheart (2012) only
found priming from truly complex words in
grade 3 and 5 English-speaking children,
indicating that morpho-orthographic priming is not
automatized yet and decomposition relies more on
semantics in developing readers. However, no
studies with children have used complex
pseudoword primes so far, although they provide
the possibility to utilize the paradigm in
languages that do not naturally have pseudocomplex
words, such as German.
Morphological decomposition in German
can be insightful to investigate, because of its
language specific characteristics. German has a
transparent orthography and is morphologically
rich. As a consequence, morphological entities
might present a very useful unit for effective
word recognition, even for beginning readers.
Nevertheless, for children being still in the
process of reading acquisition and showing more
variability in their lexical representations,
language proficiency can be expected to play an
even greater role than
Beyersmann et al. (2014)
found for adults.
The aim of the present study was
therefore to test whether the moderating effect of
language proficiency, as indexed by vocabulary and
spelling skills, on morphological priming can be
replicated with German adults and whether it
generalizes to readers at the lower end of the
proficiency range, namely children. We expect to
see evidence for a more automatized form of
morpho-orthographic decomposition in highly
proficient children (replicating the Quémart et al.
pattern), whereas low-skilled children should
show less priming
(as in Beyersmann et al.,
2012)
or no robust priming at all. In our adult
group, we expect robust priming in all three
prime conditions (including the nonsuffixed
condition) in high proficiency participants, but
reduced non-suffixed priming in low proficiency
participants
2
2.1
MethodParticipants
Twenty-four university students (13 women, Mage
= 25.2 years, age range: 20–29 years) and 24
elementary school children (13 girls, Mage = 9.5
years, age range: 8;6–10;9 years, grade 3-5)
participated in the experiment.
Each participant’s language proficiency
was assessed, using a spelling and a vocabulary
test. Adults performed a spelling recognition test,
which was modelled after the one used by
Andrews and Lo (2012). Participants were asked to
classify 100 words as correctly or incorrectly
spelled. Children performed the fill-in-the-gap
dictation test of the SLRT-II
(Moll & Landerl,
2010)
. For assessment of vocabulary, adults
completed the German version of the LexTALE
(Lemhöfer & Broersma, 2012)
, and children the
vocabulary subtest of the CFT 20 (
Weiß, 1998
).
A composite measure of spelling and vocabulary
was calculated by standardizing and averaging
the spelling and vocabulary scores for each
participant.
2.2
Materials
We conducted a masked priming lexical decision
experiment using real suffixed words
(kleidchenKLEID), suffixed pseudowords
(kleidtumKLEID), nonsuffixed pseudowords
(kleidektKLEID) and unrelated controls
(träumereiKLEID) as primes. 50 word targets were selected
from the childLex corpus
(Schroeder, Würzner,
Heister, Geyken, & Kliegl, 2014)
and 50
pseudoword targets were created by changing
one letter from a real word that was not in the
target word set. Word and nonword targets were
matched on length. For each target all four types
of primes were created. Primes were matched on
length, suffix length and non-morphemic ending
length across conditions. Four counterbalanced
lists of prime-target combinations were created,
each containing a target only once, such that
participants saw each target only with one of the
four prime conditions.
2.3
Procedure
Stimuli were presented in white 20-point Courier
New font in the center of a black screen on a 15″
laptop monitor with a refresh rate of 60 Hz. Each
trial consisted of a 500-ms fixation cross,
followed by a 500-ms forward mask of hash keys,
then a prime in lowercase for 50 ms, followed by
the target in uppercase. The target remained on
screen until response. Participants were
instructed to indicate whether the presented stimuli was
an existing German word or not by pressing a
key as quickly and as accurately as possible.
They were not informed about the presentation
of the prime.
2.4
Results
Reaction times were analyzed using linear
mixed-effects modeling. Participants and items
were included as random factors and lexical
status of the target (word, pseudoword), prime type
(suffixed word, suffixed pseudoword,
nonsuffixed pseudoword, unrelated word), age group
(adults, children) and language proficiency
(continuous measure combined of the spelling and
vocabulary scores), as well as all their
interactions, were included as fixed effects. Where
appropriate, one-sided post-hoc contrasts were
applied comparing all related priming conditions
with the unrelated condition. For contrasting
readers with higher and lower proficiency,
reaction times of participants scoring one standard
deviation above or below the mean proficiency
measure within their age group were used.
Significance was evaluated using the normal
distribution. Results are reported for word targets
only. Descriptive statistics are provided in Table 1.
For adults, priming was observed from all
three related conditions (suffixed word, suffixed
pseudoword and nonsuffixed pseudoword)
relative to the unrelated condition, z=5.04, z=4.43,
z=2.07, all p<.05. However, language
proficiency moderated priming effects. Priming in the
nonsuffixed pseudoword condition was only
significant for adults with higher language
proficiency (+1SD), z=1.74, p<.05, but not for adults
All participantsPrime TypeAdultsChildrenStimulus ExampleHigher Language Proficiency (+1SD)
For children, proficiency played an even
more pronounced role than for adults: higher
proficiency children (+1SD) showed the same
pattern as higher proficiency adults, namely
priming from all related condition, z=3.03,
z=2.02, z=2.96, all p<.05. In contrast, in lower
proficiency children (-1SD) priming in none of
the conditions reached significance, although
there was a numerical advantage from suffixed
word primes in the mean reaction times (40ms
faster compared to the unrelated condition).
3
Conclusion
Our results confirm recent evidence for French
adults
(Beyersmann et al., 2014)
, showing that
the extent to which morphological information is
exploited depends on language proficiency also
in German. Adults in the present study showed
morphological priming effects from suffixed
word primes (kleidchen-KLEID), suffixed
pseudoword primes (kleidtum-KLEID) and also
nonsuffixed pseudoword primes
(kleidektKLEID) relative to unrelated words
(träumereiKLEID). Priming from the nonsuffixed
pseudoword condition did not continue to be
significant with decreasing language proficiency.
Moreover, the pattern of priming generalizes to
beginning readers with higher language
proficiency: they show priming similar to that of
higher proficient adults. For children with lower
language proficiency, the effects did not reach
significance, but were clearly most pronounced
in the suffixed word condition.
We argue that there is a developmental
gradient in the use of morphological information
during reading acquisition, driven by language
proficiency. Beginning readers with low
language proficiency are only able to benefit from
morpho-semantic information, if at all. More
advanced lexical knowledge allows readers to
extract morpho-orthographic information.
Following
Andrews and Davis (1999)
and
Grainger
and Ziegler (2011)
, we assume that this happens
through segmentation of the affix in lower
proficiency adults, as indicated by the priming effects
of both suffixed prime conditions. Crucially,
higher proficiency adult and even child readers
with sophisticated lexical knowledge are able to
additionally use segmentation of the embedded
stem, therefore showing facilitation also in the
nonsuffixed prime condition. Our results
highlight the importance of lexical knowledge as a
further determinant of the ability to exploit
morphological structure in the word recognition
process, especially in children.
Andrews
,
S.
, &
Davis
,
M. H.
(
1999
).
Interactive activation accounts of morphological decomposition: Finding the trap in mousetrap? Brain and Language
,
68
,
355
-
361
.
Andrews
,
S.
, &
Lo
,
S.
(
2013
).
Is morphological priming stronger for transparent than opaque words? It depends on individual differences in spelling and vocabulary
.
Journal of Memory and Language
,
68
,
279
-
296
.
Beyersmann
,
E.
,
Casalis
,
S.
,
Ziegler
,
J. C.
, &
Grainger
,
J.
(
2014
).
Language proficiency and morphoorthographic segmentation
.
Psychonomic Bulletin Review.Beyersmann
,
E.
,
Castles
,
A.
, &
Coltheart
,
M.
(
2012
).
Morphological processing during visual word recognition in developing readers: Evidence from masked priming
.
The Quarterly Journal of Experimental Psychology
,
65
,
1306
-
1326
.
Grainger
,
J.
, &
Ziegler
,
J. C.
(
2011
).
A dual-route approach to orthographic processing
. Frontiers in Psychology,
45
,
Lemhöfer
,
K.
, &
Broersma
,
M.
(
2012
).
Introducing LexTALE: A quick and valid Lexical Test for Advanced Learners of English
.
Behavior Research Methods
,
44
,
325
-
343
.
Longtin
,
C. M.
, &
Meunier
,
F.
(
2015
).
Morphological decomposition in early visual word processing
.
Journal of Memory and Language
,
53
,
26
-
41
.
Moll
,
K.
, &
Landerl
,
K.
(
2010
).
SLRT-II: Lese- und Rechtschreibtest: Weiterentwicklung der Salzburger Lese- und Rechtschreibtests (SLRT): Manual
. Bern: Verlag Hans Huber.
Morris
,
J.
,
Porter
,
J. H.
,
Grainger
,
J.
, &
Holcomb
,
P. J.
(
2011
).
Effects of lexical status and morphological complexity in masked priming: An ERP study
.
Language & Cognitive Processes
,
26
,
558
-
599
.
Quémart
,
P.
,
Casalis
,
S.
, &
Colé
,
P.
(
2011
).
The role of form and meaning in the processing of written morphology: A priming study in French developing readers
.
Journal of Experimental Child Psychology
,
109
,
478
-
496
.
Rastle
,
K.
, &
Davis
,
M. H.
(
2008
).
Morphological decomposition based on the analysis of orthography
.
Language & Cognitive Processes
,
23
,
942
-
971
.
Rastle
,
K.
,
Davis
,
M. H.
, &
New
,
B.
(
2004
).
The broth in my brother's brothel: Morphoorthographic segmentation in visual word recognition
.
Psychonomic Bulletin & Review
,
11
,
1090
-
1098
.
Schroeder
,
S.
,
Würzner
, K.-M.,
Heister
,
J.
,
Geyken
,
A.
, &
Kliegl
,
R.
(
2014
).
childLex: A lexical database for German read by children. Manuscript submitted for publication
.
Max Planck Institute for Human Development
, Berlin.
Taft
,
M.
(
2003
).
Morphological representation as a correlation between from and meaning
. In E. Assink & D. Sandra (Eds.) Reading complex words (pp.
113
-
137
). Amsterdam, The Netherlands: Kluwer.
Weiß
,
R. H.
(
1998
).
Grundintelligenztest Skala 2 (CFT 20) mit Wortschatztest (WS) und Zahlenfolgentest (ZF)
. Göttingen: Hogrefe.