Sensitivity to derivational morphology as cues to lexical stress among English as second language learners

Abstract

Derivational suffixes are known to play a crucial role in assigning stress to multi-syllabic words among native English speakers. However, it is unclear whether second language (L2) learners of English can effectively use derivational suffixes as stress cues in written words. To address this gap, we studied if native Chinese-speaking adults learning English as an L2 can use derivational cues to correctly assign lexical stress, and whether the frequency of these suffixes and their L2 language proficiency moderate this sensitivity. Utilizing a written stress assignment and a stress production task, participants saw a series of written stems and their derived forms and were asked to either choose a syllable to assign the stress (Experiment 1) or read it aloud (Experiment 2). Results showed that L2 learners are sensitive to derivational cues to lexical stress in English. Moreover, the strength of this sensitivity is dependent upon the suffix class, frequency, and L2 language proficiency. Specifically, in both experiments, participants made more correct stress shifts in derived forms with high compared to low nonneutral suffix frequency; however, it did not differ between high and low neutral suffix frequency. In addition, as participants’ English proficiency increased, their correct stress shifts increased significantly greater in those derived nonwords with nonneutral than neutral suffixes in the stress assignment task. These findings make a significant contribution to the limited literature on L2 learners' sensitivity to morpho-orthographic cues to lexical stress and support the proposed statistical learning mechanism, that is, L2 learners can implicitly learn statistical regularities from linguistic materials.

Appendices

Appendix A

Stimuli used in Experiments 1 and 2

Stems	High-neutral	Low-neutral	High-nonneutral	Low-nonneutral
lescind	lescinder	lesindary	lescindity	lescindian
ralen	ralenment	ralenize	ralenic	ralenous
instean	insteaner	insteanhood	insteanic	insteanitis
ratron	ratronful	ratronary	ratronity	ratronitis
harent	harentful	harenthood	harention	harentian
fation	fationment	fationhood	fationity	fationous
linew	linewful	linewize	linewic	linewitis
vatin	vatinful	vatinary	vatinion	vatinous
halest	halester	halestary	halestity	halestian
teroic	teroicful	teroichood	teroicity	teroicitis
santed	santeder	santedhood	santedic	santedian
perimin	periminer	perminize	periminic	periminian
rerin	rerinment	rerinize	rerinion	rerinous
jated	jatedment	jatedhood	jatedion	jatedian
boohon	boohoner	boohary	boohonity	boohonous
fulwark	fulwarkment	fulwarkary	fulwarkic	fulwarkitis
tetter	tetterful	tetterize	tetterity	tetterous
difocal	difocalment	difocalize	difocalion	difocalitis
yeleep	yeleeper	yeleephood	yeleepity	yeleepian
bursitun	bursitunment	bursitunary	bursitunity	bursitunous
meblifate	meblifatement	meblifatehood	meblifation	meblifateitis
wulanna	wulannaful	wulannahood	wulannaity	wulannaitis
wibinct	wibinctful	wibinctary	wibinction	wibinctous
fulwark	fulwarkment	fulwarkary	fulwarkic	fulwarkitis
pozort	pozortful	pozorthood	pozortion	pozortitis
mozict	mozicter	mozicthood	moziction	mozictous
buvoad	buvoader	buvoadize	buvoadic	buvoadian
tetter	tetterful	tetterize	tetterity	tetterous
livup	livuper	livupize	livupic	livupian
pabect	pabectful	pabectize	pabectic	pabectous
gifial	gifialer	gifalize	gifalic	gifialian
tozorn	tozornment	tozonary	tozornity	tozornitis
nemal	nemalment	nemalary	nemalion	nemalian

The first 18 rows of stimuli were used in Experiment 2, and all the stimuli were used in Experiment 1.

Appendix B

Bigram frequency of the stems

Stems	Bigram frequency
lescind	3070
ralen	3351
instean	3670
ratron	3145
harent	3097
fation	3300
linew	3056
vatin	3928
halest	3135
teroic	3068
santed	3223
perimin	3012
rerin	3972
jated	3295
boohon	3098
fulwark	3021
tetter	3395
difocal	3011
yeleep	3218
bursitun	3341
meblifate	3067
wulanna	3290
wibinct	3870
fulwark	3756
pozort	3657
mozict	3892
buvoad	3146
tetter	3395
livup	3456
pabect	3780
gifial	3165
tozorn	3267
nemal	3786

Appendix C

Frequency of the suffixes

Condition	Suffix	Frequency (out of 1251 suffixed words)
Nonneutral-high	-ion	167
	-ity	88
	-ic	79
Nonneutral-low	-itis	2
	-ous	25
	-ian	5
Neutral-high	-er	124
	-ful	67
	-ment	43
Neutral-low	-ary	29
	-hood	3
	-ize	26

Appendix D

Mod1 < --:

glmer(change ~ frequency*class*pro + (1 + class + frequency | id) + (1|item), control = glmerControl(optimizer = "bobyqa",optCtrl = list(maxfun = 2e5)),data = morpho, family = "binomial"). mod 1 output.

	Estimate	Std. Error	z value	Pr( >|z|)
(Intercept)	0.180993	0.067509	2.681	0.00734**
frequency	0.227974	0.098289	2.319	0.02037*
class	0.831427	0.098430	8.447	< 2e−16***
proficiency	− 0.006736	0.075872	− 0.089	0.92926
frequency:class	0.632605	0.195459	3.237	0.00121**
frequency:proficiency	− 0.069788	0.100161	− 0.697	0.48595
class:proficiency	0.773274	0.100448	7.698	1.38e−14***
frequency:class:proficiency	0.115202	0.200645	0.574	0.56586

mod2 < --:

Imer (rt ~ frequency*class* proficiency + (1 + class + frequency | id) + (1|item),data = morpho).

	Estimate	Std. Error	t value	Pr( >|z|)
(Intercept)	622.731	35.284	17.649	< 2e−16***
frequency	− 31.336	14.368	− 2.181	0.0293*
class	− 26.421	14.242	− 1.855	0.0638
proficiency	28.897	30.106	0.960	0.3382
frequency:class	− 15.516	28.486	− 0.545	0.5860
frequency:proficiency	28.084	20.687	1.131	0.1306
class:proficiency	3.108	20.490	0.152	0.8794
frequency:class:proficiency	− 15.271	40.841	− 0.374	0.7085

mod2 < --:

gmer(shift ~ frequency*class*English.proficiency + (1 + class + frequency | participants) + (1 | item), control = glmerControl(optimizer = "bobyqa",optCtrl = list(maxfun = 2e5)),data = readaloud, family = "binomial").

	Estimate	Std. Error	z value	Pr( >|z|)
(Intercept)	0.04874	0.03152	1.546	0.1221
frequency	0.06072	0.08769	0.692	0.4887
class	2.18974	0.23653	9.258	< 2e−16***
proficiency	− 0.01334	0.03820	− 0.349	0.7269
frequency:class	0.31036	0.12578	2.467	0.0136*
frequency:proficiency	− 0.02480	0.10052	− 0.821	0.4122
class:proficiency	0.03145	0.28964	0.109	0.9135
frequency:class:proficiency	− 0.15135	0.15282	− 0.990	0.3220

mod4 < --:

Imer(response ~ frequency*class*English.proficiency + (1 + class + frequency | participants) + (1 | item),data = rt2).

	Estimate	Std. Error	t value	Pr( >|t|)
(Intercept)	788.878	14.938	52.811	< 2e−16 ***
frequency	− 21.020	10.028	− 2.096	0.040 *
class	− 14.441	10.936	− 1.320	0.191
proficiency	− 21.842	18.400	− 1.187	0.239
frequency:class	− 2.705	17.328	− 0.156	0.876
frequency:proficiency	− 20.435	12.304	− 1.661	0.102
class:proficiency	− 1.859	13.388	− 0.139	0.890
frequency:class:proficiency	− 28.592	21.160	− 1.351	0.177

Appendix E

See Figs. 

Fig. 1

The interaction between suffix class and frequency in the stress assignment task

1,

Fig. 2

The interaction between suffix class and English proficiency in the stress assignment task

2, and

Fig. 3

The interaction between suffix class and frequency in the stress production task

3.