This section will discuss the first studies that reported findings relevant to the main claims of the AH on emergent and developing child past morphology across different L1s, namely
Bloom et al. (1980), Shirai (1991), Shirai and Andersen (1995) for English. Several of these studies found supporting evidence in favor of some effect of lexical aspect on the use of past morphology: perfective marking occurred with achievement and accomplishment verbs and progressive meaning was marked by the past progressive at a later stage. Let us present the overall results of the aforementioned studies for a deeper treatment of the effect of lexical aspect on the use of past morphology in various L1s. However, prior to this, it is important to summarize the tenets of the AH as proposed by Shirai (1991) and Andersen & Shirai (1996):
1. Learners first use (perfective) past marking on achievements and accomplishments, eventually extending use to activities and statives.
2. In languages that encode the perfective/imperfective distinction, imperfective past appears later than perfective past, and imperfect past marking begins with statives, extending next to activities, then to accomplishments, and finally to achievements. 3. In languages that have progressive aspect, progressive marking begins with activities,
then extends to accomplishments and achievements.
4. Progressive markings are not incorrectly overextended to statives.
(Bardovi-Harlig, 2000, p. 227) Bronckart and Sinclair (1973) investigated the use of past inflectional morphology by 74 L1 French children (ages, 2;11 to 8;7). Experimentally elicited production data showed that the children tended to use present forms (present) for inherently durative events, and perfective past forms (passé compossé) for telic actions (i.e. achievement and accomplishment verbs). The role of lexical aspect on the use of past morphology decreased with development and approximated adult use. Imperfective past (imparfait) was very infrequent among the younger children emerging later than perfective past.
Antinucci and Miller (1976) studied the acquisition of past morphology through the spontaneous oral productions of one L1 English child and seven L1 Italian children and found similar results as Bronckart and Sinclair. Particularly, the participants used the past participle in Italian and simple past in English with change of state verbs with clear end results. Both studies argued in favor of a cognitive limitation in that children at an early age have not developed the concept of temporal deixis and displaced language. Furthermore, the authors stated that children used past morphology to encode the concepts that were more relevant to them, namely, events/situations with observable end results. In other words, the telicity of the situation type allowed the learners to encode such situation in the past (i.e., aspect before tense). Further support for the notion of aspect before tense can be found in Bloom et al.’s (1980) study with L1 English children. Results indicated that the emergence of inflectional forms were determined by the lexical aspect of verbs. Specifically, they argued that TA morphology was guided by the oppositions stative versus non-stative, durative versus non-durative, and completive versus non-completive. The authors concluded that the notion of aspect before tense is a relative one since although aspect seems to have a pivotal role in the acquisition of TA morphology, this does not necessarily mean that tense is relegated until the age of 6. They sustained that aspectual marking co-occurs with the learning of tense relations.
Shirai (1991), one of the most representative works in this vein, analyzed past-tense marking in the longitudinal data of three children acquiring English as an L1 (from the CHILDES database, MacWhinney, 2000). Past-tense morphology was examined in terms of the primacy of aspect hypothesis and frequency effects in the input. His results confirmed previous studies’ findings in that verb semantics was a key factor in determining the use of past morphology by the children, who used a past/perfective morpheme with telic verbs, for example. Moreover, the
attested pattern among the mothers’ input further confirmed the Distributional Bias Hypothesis in that the learner’s use of past forms showed a strong bias to prototypical combinations (e.g., perfective past and telic verbs).
The AH tenets, however, have been questioned by Bertinetto et al. (2015), who analyzed the development of TA morphology and the effect of lexical aspect among three L1 Italian children across three phases: 1. pre-morphology, characterized by rote-learned verb forms; 2.the proto- morphology stage, the moment when morphological productivity initiates; 3. modularized morphology, a stage when established morphological productivity characterizes the children’s language. The corpus data consisting of each child-caretaker interaction was analyzed so as to answer whether lexical aspect serves as a universal explanation of first language acquisition of past morphology. To this end, the data were scrutinized by means of a weighted liner regression analysis that compared the correlations of tense and grammatical aspect and lexical aspect- grammatical aspect at each developmental point. Results revealed U-shaped curves in the acquisition of past morphology consisting of a first phase characterized by a strong correlation between grammatical aspect and tense, which declined fast to increase again at a later cycle. Results also indicated an association between grammatical aspect and lexical aspect, which was stronger than the association with tense. The authors argued that this finding was consistent with the AH since perfective aspect was preferred with telic verbs and imperfect aspect was preferred with atelic verbs. A second analysis consisted of the Pointwise Mutual Information (PMI) formula, which calculated the association strength between lexical aspectual class and grammatical aspect. Results showed that stative verbs strongly favored the imperfective aspect whereas telic verbs favored the perfective aspect. Nevertheless, activities differed from stative verbs regarding the imperfect showing no polarization across grammatical aspect types like statives and telics did. The
authors discussed that it is not the semantic feature of (a)telicity that is at the heart of acquisition, for otherwise stative and activity verbs should behave in the same way. A consideration of temporality showed that the present emerged before the past and the future. The present with past reference emerged prior to perfective past in one of the participants but simultaneously in the other two. Finally, the fact that the imperfect form emerged after the past perfective allowed the authors to contend that in temporality-prominent languages (i.e., languages which mark past, present, and future contrasts overtly: Italian and German, for example) temporality contrasts are mastered prior or simultaneous to aspect. The authors proposed a typologically-oriented and morphologically- based approach to the acquisition of past morphology by which acquisition is not universally triggered by verb semantics but language-specific constraints relying on any of the ATAM categories (i.e., actionality, tense, aspect, mood).
In addition, the first empirical studies that found support for the AH in L1 morphology acquisition were rejected by Weist, Wysocka, Witkowska-Stadnik, Buczowska, and Konieczna (1984). They proposed the defective tense hypothesis, arguing that young children are not cognitively capable of envisioning events that are displaced in time and can only encode reference to the here and now. Andersen (1989a) called it the absolute defective tense hypothesis and argued that the evidence actually supports a relative defective tense hypothesis, which is descriptively identical to the Aspect Hypothesis. Weist et al.’s (1984) study was later replicated by Bloom and Harner (1989), which showed that children learning Polish were influenced by lexical aspect in acquiring verb tense. Wagner (2002) investigated the role of agency information in L1 English children’s early interpretations of TA morphology, in particular, the progressive and simple past forms. Fifty-nine children (two-, four- and five-year olds) were presented with a forced-choice sentence-to-scene matching task very similar to the one used by Weist and colleagues except that
here the scenes contained only information about the relative completion of the object of the event and no information about the state of the agent of the event. In contrast to previous research, the children here did not succeed at this object-oriented task until as late as age five; moreover, also contra previous work, when they did succeed, their performance tracked the formal entailments of grammatical aspect. Thus, subjects consistently matched the perfective sentence to the completed event (reflecting the perfective’s entailment of completion) but never consistently matched the imperfective sentence to either scene (reflecting the imperfective’s lack of entailments).
This section discussed two representative studies in favor of and against the tenets of the AH in L1 acquisition of morphology. Shirai (1991) found support for the claim that lexical aspect is a key determinant of past morphology, since children followed the main path predicted by the AH. At the same time, the children’s use of past marking and its distribution seemed to align with the tenet of the Distributional Bias Hypothesis, which holds that NSs have a tendency to associate morphology with semantically similar verb classes, and that children extract those probabilities from the input and apply them categorically to their use of morphology. Conversely, Bertinetto et al. (2015) refuted the AH in their study of L1 Italian children and their developing morphology, arguing in favor of temporality constraints and the morphological structure of the L1 as key determinants.
The following section discusses the major empirical research on the role of lexical aspect on the acquisition of past morphology in L2 English and other languages by learners from different L1 backgrounds.
2.2.1.2 Form-oriented research studies: the Aspect Hypothesis across L2 English (and other L2) instructed learners from diverse L1 backgrounds
Research on the effects of lexical aspect on the emergence and development of past-tense marking in second languages, principally on instructed L2ers, are abundant. Many of these studies have also included the role of the L1 in order to determine whether the tenets of the AH are operative across learners from different L1 backgrounds. As was stated in chapter one, one major distinction between L1 and L2 acquisition is that in the latter the learner arrives at the acquisitional process with an already established language (i.e., with an internalized grammar and semantic- conceptual features). The learner tends to learn how to map the new L2 forms to the concepts, functions, or meanings present in the L1. Thus, the learner’s cognitive maturation can be considered an advantage in terms of ease of acquisition of a second language, even though it can sometimes be obstacles in learning some aspects of the L2. This section will discuss several seminal and recent studies that investigated the effect of lexical aspect and L1 transfer, some of which support the AH tenets and others which refute them.
Robison (1990) is the first published study about the effect of lexical aspect (i.e., stative versus dynamic and punctual versus telic) on the acquisition of L2 English past morphology in a 30-year-old Salvadoran Spanish speaker, who had lived in the US for 10 years at the time of the study. An analysis of 171 predicates from spontaneous speech revealed that the participant categorically used the past form with dynamic verbs. Results also indicated that morphology depended on punctuality and telicity distinctions significantly more than on the stative-dynamic distinction (p < .001). Specific results showed that the participant used the -ing progressive past form with durative verbs and the past form with punctual verbs. Moreover, the progressive form was found across a large number of stative verbs and a couple of punctual verbs. Robison
concluded that this last pattern could indicate the participant’s shift from prototypical combinations. In sum, TA morphology use was associated with lexical aspect at least at certain stages of L2 acquisition, even if acknowledging the influence of L1, L2 and individual differences. In a later study, Robison (1995) investigated the effect of lexical aspect on the choice of past morphology in a written composition by 26 L1 Spanish Puerto Rican college students across four proficiency levels. Findings indicated that the learners tended to mark telic verbs with the past form and activity verbs with the past progressive form and these associations became stronger at higher proficiency levels. In addition, learners tended to mark achievement verbs with the past form even when they referred to present or future situations.
Bardovi-Harlig and Reynolds (1995), using a cloze test with short passages, investigated the role of lexical aspect in the acquisition of tense and aspect specifically examining the appropriateness of use between combinations of form and lexical aspect class. Their participants were college-level learners of English from a wide range of L1 backgrounds at six levels of proficiency from beginning to advanced. Results showed that the use of past with achievement and accomplishment verbs was similar across proficiency levels revealing increasing rates of appropriateness of use from Levels 2 to 6 (i.e., from 8% to 98%). Consequently, the authors grouped these verbal aspectual classes into “events” on the basis of Mourelatos’s (1981) taxonomy. The only exception to this trend was Level 1, which obtained 73% appropriateness with accomplishments and 62.4% accuracy with achievement verbs. The accuracy rate of simple past with activity verbs remained low and only obtained 80% appropriate use in the highest proficiency level (i.e., Level 6). These results provided support for the tenets of the AH regarding the simple past form in prototypical combinations at beginning stages and the appropriate use of non- prototypicality at advanced proficiency. In terms of developmental sequences, results showed that
learners go through three distinct stages in the simple past acquisitional process: 1. telic verbs appear before atelic verbs; 2. states seemed to be used more than activities; and 3. the use of the simple past appears to be undergeneralized.
Housen (2000) studied the development of TA morphology in one L2 English learner (Ema) and found evidence against the role of verb semantics on past marking as an absolute acquisition universal. He claims that lexical aspect constrains morphology, but it is itself constrained by other factors, such as L1-induced predispositions to mark specific temporal categories, the morphophonemic nature of different grammatical categories, and the nature of the processing mechanisms in the learning of grammatical morphology. In particular, the data revealed that in line with the AH, Ema’s past marking at Time 1 strongly emerged with regular telic verbs, reaching the highest association at Time 2, at which point the association relaxed. However, perfective marking was also highly associated with states (i.e., against the AH), which were realized by highly frequent irregular verbs in the input. Housen explained the dissociation of regular and irregular morphology by referencing the processing Dual-Mechanism (Pinker & Prince, 1994). This model suggests that irregular morphology is learned through associative memory or rote-learning, whereas regular morphology is acquired through rule-learning. Irregular morphology emerges first due to the verbs’ high frequency`, which is the reason for their entrenchment and conservation in language.
A cross-sectional study by Ayoun and Salaberry (2008), which analyzed data from a group of 21 high-school L1 French speakers learning English as a foreign language in France in a formal setting. The purpose of this study was to find evidence that supported the effect of both the AH and L1 transfer in the acquisition of past tense morphology in English. The participants participated in two written elicitation tasks: a personal narrative and a cloze task. For the former
task, participants were asked to write a personal narrative or a fairy tale; for the latter, they were instructed to fill in the blanks with an appropriate tense, given the base form of the verb in parentheses. The blanks mostly included the simple past contexts across the four lexical aspectual classes of verbs. The results of both elicitation tasks revealed a strong lexical aspect effect associated with the use of past tense markers in L2 English. With regard to the cloze task, results showed that learners achieved high consistency scores for both stative and telic predicates. These findings emphasize the effect of lexical aspect on the use of past tense markers, while suggesting a significant departure from the predicted developmental path of past tense marking: states are marked more consistently than telic events in the narrative. These findings are explained in light of task effects.
Upor (2009) examined the development of past morphology and the effects of lexical aspect and L1 influence among L1 Tanzanian learners of L2 English in multilingual settings. The participants attended primary school, secondary school, and university and they varied in terms of proficiency level from beginner to advanced. Participants spoke one Bantu language, which morphologically marks tense and aspect. As an agglutinative language, Bantu has a verb system with several morphemes that mark other grammatical categories. A picture-story-retell and a personal narrative constituted the elicitation tasks used and the major finding was partial adherence to the AH. Particularly, most groups initially marked the past primarily on statives and then on achievements, a result that yielded statistical significance. Moreover, when the production of achievements outnumbered that of the stative verbs, past marking occurred on telic verbs. Support for the tenets of the AH was observed with the progressive, which emerged with activities and spread gradually to the other predicted verb classes. The author concludes that lexical aspect may
represent an incomplete explanation of the development of EFL among L1 speakers of a Bantu language in an instructed setting.
Chivarate (2018) analyzed the effects of verb semantics and L1 background on the acquisition of the simple past and the past progressive in 5 proficiency-level groups of Thai EFL learners. Thai is typologically different from English exhibiting neither morphological encoding of tense nor aspect. Thai makes use of pragmatic devices (e.g., chronological order in narration) and lexical means (e.g., adverbs such as yesterday, last month, the next year) in order to refer to temporal points. It has a number of grammatical aspect markers to express the different viewpoints with regard to events. Cloze test results revealed that as L2 proficiency increased, the learner used past morphology more accurately. A comparison within a proficiency group showed that learners use simple past morphology more appropriately than past progressive morphology. Lexical aspect was also found to have an effect on past marking: in terms of telicity, the simple past was preferred with telic verbs (i.e., achievements and accomplishments) and in terms of the Vendlerian classes, the form was greater with the [+punctual] predicates (i.e., achievement) than the [-punctual] predicates (i.e. accomplishment, activity and state). In contrast, the use of past progressive morphology was greater with the [-punctual] predicates (i.e. accomplishment and activity) than with the [+punctual] predicates and regarding telicity, past progressive morphology was greater with the [-telic] predicates (i.e., activity) than the [+telic] predicates (i.e., achievement and accomplishment). Overall, the prototypical combinations obtained a higher rate of appropriate use of past morphology. The results were consistent with the AH and in agreement with findings from Robison (1995) and Bardovi-Harlig and Reynolds (1995). Finally, the learner’s IL showed different L1-influenced forms, suggesting that L2 morphology development is conditioned to a lesser or greater extent by the L1. For example, Thai expresses both definite and indefinite pasts
with the tense-aspect marker laew, whereas English does so by means of the simple past-present perfect dyad. This typological difference may explain the low proficiency learner’s overuse of PP in definite contexts, where the simple past was required. The author argues that this last result supports previous L1 transfer studies (e.g., Collins, 2002), which also found an overuse of the perfect among her L1 French learners of English but of higher proficiency level than the ones in Chivarate’s study. The conclusion is that L2 development is conditioned by L1 influence.
Zhao and Shirai (2018) recently investigated the acquisition of past morphology and the role of lexical aspect by Arabic-speaking learners of English at beginner and intermediate- advanced levels.12 The AH was examined on the basis of the accuracy of past tense marking by lexical aspectual class calculated on the basis of Suppliance in Obligatory Contexts. In general,