Fossilization of English as an L2 non-native constructions at the syntax-

discourse interface in Romance L1 speakers

Fosilización de construcciones no nativas del inglés como L2 en la interfaz

sintaxis-discurso en hablantes de L1 románica

Marciano Escutia López

Universidad Complutense de Madrid

mescutia@ucm.es

ABSTRACT

This study examines and compares the English as a foreign language production of two groups of

speakers with respect to certain manifestations of the expression of the subject in discourse. One

of the groups consists of 12 advanced EFL university English majors living in Spain, the other of 9

adult Romance speakers having lived in an English speaking country for several decades. The

hypothesis is tested that subject inversion with unaccusative verbs and pronominal subject omission

when identified in discourse constitute fossilization phenomena as verified by their perseverance in

the second group of learners despite their long and rich exposure to the L2 and their frequent

interaction in it.

Keywords: English as a second/foreign language, (non) null subject language, fossilization, syntax/discourse

interface, adult language learning

RESUMEN

Este estudio examina y compara la producción del inglés como lengua extranjera de dos grupos de

hablantes con respecto a algunas manifestaciones de la expresión del sujeto en el discurso. El

primero es de 12 alumnos españoles avanzados de inglés como lengua extranjera al final de su

licenciatura en Filología Inglesa, y el segundo de 9 hablantes adultos de lenguas románicas

residentes en países de habla inglesa durante décadas. Se intenta comprobar la hipótesis de que la

inversión verbo-sujeto con verbos inacusativos y la omisión de los sujetos pronominales

identificados por el discurso en la lengua nativa constituyen elementos de fosilización si su

pervivencia se mantiene en el grupo segundo, a pesar de su larga y abundante exposición a la L2 y

su frecuente interacción en la misma.

Palabras clave: inglés como lengua extranjera/segunda, lenguas de sujeto nulo, fosilización, interfaz

sintaxis/discurso, aprendizaje adulto de la lengua

Revista Nebrija de Lingüística Aplicada a la Enseñanza de Lenguas (RNAEL) ISSN 1699-6569

Vol. 15 Núm. 30 (2021) doi: 10.26378/rnlael1530408

Recibido: 10/03/2020 / Aprobado: 10/01/2021

Publicado bajo licencia de Creative Commons Reconocimiento Sin Obra Derivada 4.0 Internacional

1. INTRODUCTION

The goal of this paper is to make the case for the fossilization of certain non-native

constructions of English as a second and foreign language (EL2, EFL, respectively. (2They

are related here to the expression of the subject by adult advanced Spanish learners

(mainly) as well as by other Romance language speakers. In order to do that, we will first

describe the concept of fossilization in L2 very briefly and concisely, its conditions and

possible causes. Then we will refer to some studies on fossilization about both the EFL of

advanced romance language students and of a romance language as a FL by advanced

English speaking students. Both are related to very similar manifestations of fossilization

dealt with here. Next, our data and hypotheses will be presented, comprising written and

spoken EFL/EL2 production of advanced adult romance speakers from both tutored and

untutored milieus, some living in Spain and others who have been living in an English-

speaking country for a long time but all of them having really started learning and/or being

exposed to the language in post-adolescence. They are presented to witness to our general

hypothesis that the kind of items referred to are true examples of fossilization. Finally, our

data will be discussed and some very brief comments will be made regarding how they

may impinge on the learning and teaching of the overt manifestations of the subject in

EFL.

Neither the linguistic features, the psycholinguistic mechanisms nor other kinds of

factors involved in L2 learning will be the central issue here since our interest is purely

factual, in the product, very concretely oriented and not directly concerned with the

specifics of the process. Our aim is not so much to look at the L2 knowledge of the learners

as to point to some particular linguistic elements and discourse constructions as candidates

for fossilization. We look at the English of advanced L1 Romance-speaking adult learners

and compare the linguistic production of two samples of them, one from last year L1

Spanish college English majors and the other from learners who have had many more

years of exposure and practice in the L2. The non-native constructions pointed out will be

referred to specific manifestations of an underlying linguistic description and some

comments will bear on the characteristics of the linguistic product and its acquisition.

This is a qualitative study so a quantitative analysis will not be carried out of the possible

contexts of actual native or non-native-like items with respect to a particular construction

from the production of these L2 learners (E2Ls. (2We just want to verify that the types of

items presented occur normally in the linguistic production of both groups in order to high-

light such “inconsistency” between native and non-native-like constructions present in the

L2 production. Furthermore, it is not our concern here whether these non-native

constructions only affect the level of externalization of the language or if they concern the

underlying L2 competence.

2. WHAT IS FOSSILIZATION

Selinker (1972) first called fossilization one of the features of L2 production whereby a

particular kind of (non-native) error1 seems to persevere without wholly dissapearing

except in highly monitored tasks. This pervasive property of interlanguage (IL), or

transitional system of the L2 created by the learner, does not appear to be permeable, at

least in production, to either tutored or untutored abundant input, a fact which he ascribed

to neurological factors hindering the complete and permanent learning of certain structures

of the L2.

Lightbown (1985) characterized it as the natural end of many aspects of the grammar

of adult L2 learners and Zobl (1980) hypothesized that fossilization is the result of

crosslinguitic influence in IL creation, which delays learning. Gass (1997) in her Input-

Output model of L2L mentions that fossilization occurs when input does not succeed in

restructuring the L2 learner’s linguistic system.

Summarizing with Savillle-Troike (2006), the term fossilization for L2Ls refers to the

fact that many of them will stop their IL development in some areas before reaching target

language norms. She points out as key factors in the process age, social identity,

communicative need and lack of motivation.

Han (2004, 2013) in her state-of-the art reviews on the topic and gathering information

from abundant studies and other summaries like Lightbown (2000, 2003) and Han & Odlin

(2006), states that in order to properly talk about fossilization certain preconditions have

to be met, like having the appropriate motivation to learn the language and getting both

abundant exposure and enough interactional practice. She also remarks that fossilization

is a local and selective phenomenon, affecting only certain individual elements of the

grammar, not the whole of it; that it truly becomes evident when the L2Ls use the language

to express their own meanings; and that L1 transfer is the major influence encompassing

it along with age-related maturational constraints. She also notes that fossilization belongs

in adult L2 learning rather than children’s, and that the following forces conspire towards

facilitating it: (a) the variable nature of the target structure (for example, an L2 apparently

displaying optionality between both presence or absence of S-V inversion or null subjects);

(b) non-robust input, because of a dearth of frequent contact with L2 native speakers and

real discourse; (c) crosslinguistic influence from an L1 unmarked usage, which may

happen, for example, when the L2 fossilizable structure has a wider distribution than in

the L1 (again, for example, the apparently optional presence of S-V inversion or null

subjects in L2 Spanish with respect to L1 English. (2Finally, she defends that fossilization

ought to be studied longitudinally in order to verify a well-established lack of progress in

those linguistic features supposed to have stopped in their development. All these are the

elements of her Selective Fossilization Hypothesis (2009. (2Thus, her notion about the

phenomenon could be summarized as a stable want of control of a second language

(feature) despite constant exposure to robust input.

In terms of the psycholinguistic factors involved in the process, Long (2003) adds that

the fossilization of a particular feature occurs because of lack of sensitivity to noticing the

difference between the L2 input and output. This, in turn, he notes, is proportional to the

perceptual saliency of the construction itself.

3. FOSSILIZATION IN ENGLISH L1/L2 AND ROMANCE L1/L2 LEARNING

The types of items discussed here are constructions found in the interaction of a Romance

L1 with EFL/EL2. They could be said to belong to what Sorace (2004) calls “soft syntax” or

grammatical aspects at the syntax-discourse interface, as these are L2Ls’ non-native

constructions occurring within a text, either written (as in the case of our group of

advanced college students) or oral (as in the interviews and lectures of our group of adults

having lived many years in an English-speaking country. (2She explains that they are not

purely formal non-native constructions, are intrinsically very hard to acquire and appear

late developmentally (Sorace & Keller 2005. (2This is so because of discourse aspects,

which might only be factored in if there has been a lot of rich environmental L2 input

exposure. Furthermore, adult L2Ls most probably behave here differently from child

acquirers because, unlike the latter, whose knowledge of the correspondences between

grammatical forms and their functions is acquired without cross-linguistic influence, adult

L2Ls might be experiencing a competition of other factors as well, mainly related to their

L1 (Sorace 2004, 2005).

We will examine briefly some of the research that has been carried out in L2L with

respect to the expression of the subject. Though we distinguish between those studies that

have concentrated on the (external) syntax-discourse interface level from those that have

done so on the (internal) syntax-semantics one, things are not so clear-cut. It may be

simply a matter of where they have placed the weight of the linguistic computation giving

rise to the non-native constructions of their studies.

3.1 Research at the syntax-discourse interface level

Research in this area, at the level of the externalization of grammar and its interaction

with pragmatic-discourse factors in production, points to attrition in the knowledge of the

L2 relation between morpho-syntactic properties and discourse pragmatics or form-

meaning-function (FMF) mapping (Lardiere 2006, Montrul & Bowles 2009, Sorace &

Serratrice 2009, Han 2011; Han & Lew 2012. (2Besides, some of that research concludes

that fossilization may extend also to some structures at the internal syntax–semantics

interface (Montrul & Bowles 2009, Sorace & Serratrice 2009. (2

One of the phenomena that has been studied in Spanish L2 by L1 English speakers is

unaccusativity, which will also figure here in our own data. This is the property of certain

intransitive verbs whose internal NP theme argument is the syntactic surface subject

requiring inversion in Spanish according to contextual conditions. On the other hand, in

standard English, inversion requires the preverbal subject slot to be filled with the

pleonastic pronoun there2, e. g. Why did you send for more food? Because there came

many more people in the end (rather formal)/many more people came in the end vs

(Spa) porque (ø) vino mucha más gente al final/??porque mucha más gente vino al

final. The interrogation marks show that though in Spanish S-V inversion might seem to

be optional in contextless syntactic terms, when the subject represents new information in

focus, inversion is required with such verbs3. Despite the fact that advanced adult L2Ls of

Spanish know rather early that both preverbal and postverbal subjects are syntactically

possible in Spanish they do not use them properly in discourse (cf., for example, Liceras

1989; Liceras & Díaz 1999; Al Kasey and Pérez-Leroux 1998; Hertel 2003; Lozano 2002,

2006; Montrul and Rodríguez Louro 2006; Dominguez, L. & Arche, M. J. 2014. (2In the

case of tutored contexts, this is an aspect hardly practiced in class. The corresponding

construction is, thus, amenable to fossilization, or as Sorace (2005) calls it, “residual

optionality”, in the sense that some specific manifestation of it is implicitly considered

optional and there might lie the lack of ultimate attainment in proficiency.

Another area that has been the subject of studies about fossilization is the distribution

of null and overt subjects in romance languages: pronoun subjects may be null in null-

subject languages when referring to something already mentioned or understood in

context; overt ones are used when they refer to new participants in the predication or for

contrast with others (Fernández-Soriano 1991)4. Advanced adult L2 learners from non-null

subject L1s soon learn that both null and overt subjects are possible in Spanish, but they

find it difficult to use them appropriately in discourse production, ruled as they are by those

FMF properties of the syntax-discourse interface (see, for example, Pérez-Leroux and Glass

1999; Liceras and Díaz 1999; Lozano 2003, 2006, 2009; Hertel 2003; Montrul 2005, 2006;

Sorace and Filiaci 2006; Belletti, Bennati and Sorace 2007, Pladevall 2013. (2

Sorace (2005) found that near-native speakers of Italian overgeneralize overt subject

pronouns and preverbal subjects to contexts which would require null subjects and

postverbal subjects in native Italian. Obviously, these advanced speakers do not lack

syntactic knowledge and when they use null pronouns and postverbal subjects, they do so

correctly, that is, they have acquired a null subject grammar as shown in context-free

grammatical tasks. Again, it seems that the appropriate FMF mapping conditions for the

use of overt and preverbal subjects have not been learned as they present optionality

effects in their use. These may be due to the lack of specification of the interface features

[topic-shift] and [focus], as Sorace points out, which prevent the speakers from

interpreting overt subjects as shifted topics foci.

There is also research about residual optionality changes in the pronominal system of

native Italian speakers who have had very extensive exposure to English (Sorace 2000;

Tsimpli, Sorace, Heycock and Filiaci, 2004. (2Its results seem to show that these speakers

display a similar pattern of optionality as the English near-native speakers of Italian, since

they appear to extend overt subjects and preverbal subjects to contexts requiring a null

subject or a postverbal subject, both in production and in comprehension tasks. Thus,

there is an overlap between the end-state knowledge of English near native speakers of

Italian and the native knowledge of Italian near-native speakers of English with respect to

null/overt subjects and pre/postverbal subjects. In both cases, the speakers’ grammar

specifications licensing null subjects are in place but the syntax-discourse mapping rules

on pronominal subjects are affected by attrition. However, this occurs for different reasons:

in the case of the former group probably due to the lack of robust L2 data in their

environment and in the latter, just the opposite, because of the length of their exposure

to such robust input of contextualized native English, which has ended up permeating the

L1 with some of its interface conditions.

3.2 Research at the syntactic competence level

Although the studies reviewed so far restrict fossilization to the syntax-discourse interface

leaving syntactic knowledge intact, other studies suggest that this attrition might also

affect syntactic competence as well. These are studies that have used both oral or written

production tasks and grammaticality judgements, which might seem to be more directed

to the learners’ purely syntactic knowledge since no context is provided.

Lardiere (2007) presented a native speaker of Chinese, Patty, who had lived as an adult

in the USA for more than 20 years, with two grammaticality judgement tasks on English

adverb placement 18 months apart. She found that Patty’s knowledge of verb raising, as

related to adverb placement (for example, with respect to examples such as *the chef

cooked slowly the meat), was native-like but also that her verbal regular morphology

across both tasks might be considered fossilized (as compared with irregular verbs. (2The

findings of this study confirmed the results of a previous one looking only at production

tasks (Lardiere 1998. (2

The same attrition symptons have been found in EFL near-native Spanish speakers with

respect to that-trace effects in grammaticality judgments, fill-in-the blanks and wh-

question formulation tasks (Escutia 1999. (2The presence of the complementizer that in

such construction (e.g., Who did you say *that came to the party?) seemed to be optional

in the grammar of these advanced speakers, especially in the case of those whose real

exposure to the L2 had begun in their post-adolescent years. Most of them were EFL

teachers themselves who had never been made aware of this constraint. This would seem

to be an example of residual optionality in their grammatical representation since no

discourse to respond to was provided.

In Escutia (1998, 2002) two study cases were reported of two respective adult Spanish

advanced learners of English who had both lived in an English speaking country for a long

time. Their L2 spontaneous (everyday conversation) and semi-spontaneous oral

performance (production uttered during a recorded interview) was gathered and presented

later to them in disguised form as individual contextless items in an acceptability judgment

task. Some (then considered) derived properties of the Pro-Drop and Verbal Agreement

parameters of Universal Grammar (UG), or innate constraints in the development of

natural grammars, were examined trying to show that the learners had not fixed them

univocally. The studies also tried to show that their L2 competence -as seen both from the

two different tasks, one more introspective and the other more spontaneous- was rather

uniform in that both their judgments and oral production showed lack of detection of some

of the differences between the L2 native grammar and their own non-native one.

Thus, the knowledge shown by the oral production data could not solely be ascribable

to circumstantial performance or discourse interface factors but reflected some kind of

more stable underlying grammatical representation. It was concluded that both learners

had developed an L2 grammar where certain aspects, supposedly derived from the fixation

of parameters, had not been univocally acquired. This seemed to be the case as indicated,

specifically, by the following phenomena: the presence of some null pronominal referential

subjects (those in embedded subject-correferential clauses: They don’t think that *(they)

are coming); the absence of some pleonastic pronouns (especially in embedded clauses: I

think *(it) is important to be there); inverted subjects with unaccusative and passive verbs

(but then (it) happened many things; and also it was established a school) and in

embedded relative clauses with a relativized object (The moment that represents the

picture is my graduation); the presence of that-trace effects; and adjacency violations or

adverbial interruptions between a verb and its object because of L1 transfer of verb raising

(I saw physically the university. (2Still, in spite of the convergence found in the two types

of tasks, the more introspective one was somewhat more accurate probably because it

taps closer to the syntactic competence of the speakers while the other, less accurate, has

to deal more with discourse interface conditions.

On the other hand, Escutia (2008, 2010) studies how English and Spanish as a foreign

language (EFL and SFL, respectively) advanced students seem to construct their L2

grammar with respect to unaccusative and passive predicates as seen in consistent written

production data. As mentioned before, these are intransitive predicates which favor verb -

subject inversion (both in Spanish and in English) and (overt) expletive anticipation of the

notional postverbal subject (in Standard English, there for NPs and it for clausal subjects.

(2Those data, in turn, are examined in Escutia (2012) to compare the structures produced

by those learners, and conclude that both types of advanced learners may be using

similarly expletive or default pleonastic it and the Spanish pronoun se, respectively,

producing parallel fossilizable structures (e. g.: *it happened something terrible / *se

ocurrió algo terrible. (2The fact that Escutia (2016) also finds converging introspective

data by an advanced SpL2 learner with very long and rich exposure to the L2 unaware of

judging his own naturalistic production with se overgeneralizations as individual

contextless items (e. g. *se faltaron muchos a clase) might suggest that it and se could

have equivalent syntactic value in their respective L2 grammars in constructions that are

subject to fossilization.

Moskovsky and Ratcheva (2014) examined two-year longitudinal data from a Russian

university teacher, an advanced fluent adult learner of English, with excellent cognitive

and motivational dispositions, who had been living in an English-speaking country for eight

years totally inmersed in the L2 language and culture. He showed symptoms of fossilization

in his production with respect to different types of items but the study centered on article

use in particular, looking at his fluctuation between native and non-native usage and

backsliding. They also presented their learner with a grammaticality judgement task to

test his intuitions about the use of articles in English, which was also presented to eight

native controls, showing a clear difference in knowledge with the latter. Thus, these

authors suggest that fossilization is a competence phenomenon, rather than just a

performance one, when the learners are trying to produce their own meaning.

The results of the previous studies, then, are congruent with Han’s (2006) remarks

about the consistency between grammaticality judgments and naturalistic data. In fact,

she concludes that grammaticality judgments may well be a “viable alternative for studying

fossilization” (p. 76. (2Thus, such a more monitorized type of task, may confirm prior

findings based on naturalistic production within ongoing longitudinal investigations of

fossilization.

Nevertheless, current research (see, for example, Lozano & Mendikoetxea 2013 and

Mendikoetxea & Lozano 2018) centers at the same time on the interface between syntax

and externalization (both discourse and phonetic realization) and on the language-internal

interface (lexicon-syntax. (2Authors favor, thus, a multi-interface approach in order to

provide a deeper understanding of the factors involved in, for example, inversion in L2

acquisition and in interlanguage grammars in general by using corpus and experimental

data, looking both at core syntax and the interfaces and considering representational and

processing models as well.

In any case, whether fossilization is a competence or a performance phenomenon, or

has to be approached from a more global perspective, taking into account both internal

and external interfaces, our goal here is just detecting it both in the oral and (more

monitored) written production of certain constructions by adult learners of EFL/L2. In order

to do that, it is not necessary to go into either their theoretical linguistic underpinnings,

the locus of occurrence or source of their attrition.

3. THE STUDY

4.1 Learners and types of items

As mentioned in the introduction, the data for this study have been drawn from two

sources: first, in the case of the twelve linguistic majors, from their written assignment

papers about applied linguistics; second, in the case of the nine Romance language

speaking personalities, from oral interviews (mainly) and a few video lectures. The former

are last year college students with an advanced level of English as measured by their

having passed the C1 level exam of the Common European Framework of Reference5 the

year before. The latter group is made up of nine learners: four speakers of peninsular

Spanish, one of Mexican Spanish, one of Brazilian Portuguese and two of Italian. Seven of

them have spent more than twenty years in English speaking countries having to use the

L2 in their jobs and are still living in those countries for most of the year. Although the

other two do not live in English-speaking countries, they have acted in many English-

speaking films and spent considerable time both in England and the United States. Despite

not all having the same L1, L2Ls from typologically similar L1 backgrounds tend to fossilize

around the same linguistic elements (Trenkic 2009; Balcom 1997; Oshita 2001. (2They

have been chosen because their production data are publicly available and because they

started really learning English in their post adolescent years, as did the college students,

most of whom just took an EFL compulsory class every year of their pre-university school

life.

Both groups are trying to get their own meaning across but through different means and

different amounts of planning because the students’ papers do require more and, in that

sense, correspond to a more monitored kind of production than the rather spontaneous

one of the oral interviews or even the lectures of one of the personalities, which are not

read but delivered orally using powerpoint displays. If the types of non-native constructions

examined also occur in the personalities’ less planned L2 production, this would support

their not being idiosincratic performance mistakes but rather resulting from similar

underlying linguistic specifications and their interaction with contextual factors. Showing

that their non-native constructions are similar to those of our advanced students would

point to their being part of a set that tends to fossilize. Let us not forget that fossilization

should be most evident in production tasks like these, where learners primarily attend to

meaning and draw from their own linguistic resources (Ellis 2003: 16).

The fact that we are contrasting data from two clearly diverse groups, one from a tutored

academic L2 learning setting and the other from an untutored and more unstructured one

does not constitute a drawback of the study, just the opposite: if it is found that both

groups produce the same kind of utterances, which is our general hypothesis, these may

be candidates for fossilization since they occur both in rather spontaneous production as

well as in a more introspective kind, as the students do not have to encode the L2 on the

spur of the moment, unlike the personalities, but can think more about it. Thus, if that is

the case, their presence might signal their rather persevering status in the IL of adult L2Ls

with the same or typologically similar L1s. In fact, if speakers of typologically similar L1s

coincide in the type of error, it may be a sign that the factors at play are mainly linguistic

and not idiosincratic. Furthermore, although both groups are of adult learners, the

academic group is rather younger than the other. This plays in favor of the candidacy of

fossilization for the tested items since the long passing of time with exposure to robust

input of learners with good cognitive capacity and motivation (and continous study of the

L2 in most cases) does not seem to change the outcome for the second group, again, a

sign of attrition in L2L.

As mentioned above, fossilization is a local phenomenon (cf. Han & Odlin 2006), and,

as such, only some specific units of fossilization related to the expression of subjects will

be examined, not all the possible manifestations of a supposedly native overarching

linguistic setting encompassing them. After having observed their pervasiveness in the

production of high-intermediate and advanced EFL students, the non-native items chosen

to confirm their condition as possible fossilization candidates and whose presence in both

sets of data we predict are the following: (1) null expletives and null subordinate subjects

correferential with a main or matrix clause referent: I think *(it) is important to be there;

They don’t think that *(they) are coming (cf. White 1985, Liceras 1989, Phinney 1989

Escutia 1998, 2002 and many others), presented in two different chart columns in the

appendices, and (2) inverted subjects -both clausal and phrasal- with unaccusative and

copula/passive verbs anticipated or not preverbally by a non-standard inserted expletive:

(it) happened many things; it was established a school6 (cf. Escutia 2008), again

corresponding to two different columns. Because of our previous experience with students,

we should expect inversion more frequently anticipated by the expletive than not.

The frequent production of both types of items already found in the studies mentioned

above, suggests a lack of command of the distributional properties of the target

construction in advanced EFL learners. English requires overt expression of the subject of

a clause, whether main or subordinate (except for imperatives and some minor clauses (as

in (∅) Stop it! As soon as (∅) possible; (∅) Born to poor parents, he still succeeded in life.

(2Spanish, on the other hand, allows both null and overt subjects, depending on discourse

pragmatics. The variable nature of subject marking in Spanish, which does not favor overt

expression of its pronominal subjects once mentioned or recoverable from context, as it

would signal another referent (Fernández-Soriano 1991), constitutes a super-set with

respect to English, where there is only one possibility no matter what the discourse

conditions are (e.g. Cuando Pedroi llegó a casa, éli??/j empezó a comer ‘when Pedro came

home, *(he) began to eat’).

The same happens with subject-verb inversion, which is freer in Spanish though it is

required syntactically with unaccusative and reflexive-passive verbs (which are really much

more frequent than periphrastic passives: Se abrieron las puertas > las puertas fueron

abiertas: ‘seref opened the doors’ > ‘the doors were opened’. (2Inversion is preferable

with intransitive verbs as the focus position of new information without, as in English, an

anticipatory expletive of the post-verbal subject (e. g. ¿Qué ocurrió? ‘What happened?’:

(1) Ladraron/estuvieron ladrando los perros toda la noche ‘*barked/were barking the dogs

the whole night’: the dogs barked/were barking the dogs the whole night; (2) Sonaron

muchos timbres toda la noche ‘*rang many bells/ many bells rang/there rang many bells

the whole night. (2Again, English here is a subset of Spanish, where the subject can

syntactically occur both pre and post-verbally though the latter option is preferred for

informational focus.

The two items chosen here as predicted to occur in both sets of data, then, production

of inverted and null subjects, might seem to have to do more with accuracy at the syntax-

discourse interface because, in terms of semantic alignment, the L1 and L2 are congruent

with respect to the syntax-semantics interface, as the subject in the case of unaccusatives

and passives in either is really the internal argument of the verb. Besides, both the subject,

in the case of the expletives in English, or the null pronouns in Spanish have no semantic

role or, having one, it is discoursively recoverable, respectively.

4.2 The data

Because this is a qualitative study, only the error tokens will be provided, in order to show

if their type of item forms part of these learners’ IL. We have also observed that the

corresponding native-like counterparts (with non-inverted subjects and overt pronominal

ones) also belong in their IL, as they are advanced proficient L2Ls. It is precisely this

residual optionality having to do with discourse factors that characterizes fossilization.

4.2.1 The students’ written data

As explained above, these data correspond to non-native constructions from the written

essays of 12 college students about different topics in Applied Linguistics. Nine of them

were female, corresponding more or less to the female/male ratio of students obtaining in

our English department. We think that this female predominance in the sample is not a

drawback of the study since it has been shown that women have, in general, higher

linguistic ability than men. What is meant is that, apart from other motivational or social

factors, when learning language, girls’ brains show greater activity in the areas used for

language encoding while boys’ show activity in the areas associated with visual and aural

functions (Burman, Bitan & Booth 2008. (2This may play in favor of the production data

being in fact more native-like. All of them were done at home and under the express

proviso of using their own language, not copying from texts, as it is evident from the non-

native constructions they show.

Table 1 is provided below with the total number of tokens per type of item. Appendix

I shows the specific utterances the students produced. Although, for reasons of space, the

broad context of the utterances is not given one can easily ascertain the topic being

discussed if one is minimally acquainted with certain areas of Applied Linguistics.

Item

Student

Stu

III

Stu

VII

Stu

VIII

Stu

XII

Inversion

2,4

2,It +

inversion

4,5

Null S

Main

1.08

Null S

Subord

1,16

Table 1. Students’ results

4.2.2 Personalities’ oral production

As already mentioned, nine L1 Romance-language personalities have been chosen as a

second group of L2Ls for comparison: four male speakers of peninsular Spanish and five

female ones (one speaker of Peninsular Spanish, one of Mexican Spanish, one of Brazilian

Portuguese and two of Italian. (2All of them are, then, speakers of typologically similar

languages in terms of subject omission and inversion7. The personalities have been chosen

-apart from the reasons given before- because it is not easy to find other type of people

with the same conditions of exposure, length of residence and abundance of data. The

names that will be used are fictional ones in order to keep their anonymity as safe as

possible8.

Although they have their own different backgrounds, all of them have in common that,

at some point in their careers, they took the step of going to work and/or live in an English-

speaking country either for life (Jess and Gaby), or as their operational base where they

spend long periods of time. In all cases they have been learning the L2 for at least twenty

years (most of them many more), have had language coaches and have taken formal

classes for long periods of time. They do not use it in the same degree because in the case

of three of them their partners are L1 speakers of the same language, but all have to use

and be exposed to it much of the time for their work and social life.

Let us provide some more detail of their English language backgrounds. Jess, 86, has

lived in the United States for almost sixty years since he arrived in his late twenties

knowing little English. His data come from both oral lectures and interviews. The case of

Alda is a bit different from the others because she does not use English so much nowadays

but has done so very often for most of her professional life and frequently gives interviews

in the language. She is now 86 and went to work in the USA in her early twenties, lived

for long periods of time in English speaking countries and has interacted in the language

very often. Her case is like that of Yoc, 77, in their continuous exposure and interaction in

English for decades. Maddy, 56, studied some EFL in school but when she started her

professional life in the USA in the early 1990s, she could speak very little and, according

to herself, she just picked it naturally because she has a good ear for languages

(reportedly, she speaks three foreign languages, apart from his native Italian) but she has

always used a language coach for her many films in English. Gaby, 70, is Brazilian and

went to live and work in the United States in her mid-thirties knowing little English and

has lived there since then. Both Raúl, 51, and Roz, 60, had to learn the language at the

workplace -like Yoc- in their early twenties in the USA, where, at the beginning, they had

to read their lines phonetically by heart though all three would later take formal lessons

for years. Bea and Belle, 46 and 51 years old, respectively, started learning the L2 in their

late teen years (apart from the EFL classes they had in high school) and even lived in the

United States for one year as students before returning to their countries first and then

going back again two years later, after which they have spent most of the time there

working in English. This earlier exposure to English shows both in their fluency and

pronunciation as compared with the others -except, perhaps less clearly, with Gaby and

Maddy- who are as fluent but have a bit more of a foreign accent.

One may safely say that the data of all these speakers probably reflect the end-state

grammar of mature and highly-exposed-to-the-language seasoned L2Ls, which may

provide a window into aspects of ultimate attainment and fossilization. The learners’

suitability for participation in fossilization research is optimal in terms of length of

residence, cognitive capacity, learning conditions, motivation, exposure to input and

opportunity to engage in authentic communication (Han 2013).

Table 2 below displays the total number of tokens per type of item. For each speaker, a

table is provided in Appendix II with their respective oral utterances.

Item

type

Yoc

(pen.

Sp.)

Raul

(pen.

Sp.)

Jess

(pen.

Sp.)

Gaby

(brazil)

Bea

(Mexican)

Roz

(pen.

Sp.)

Belle

(pen.

Sp.)

Alda

(Italian)

Maddy

(Italian)

Inversion

1.5

It+invers

3.5

Null S

Main

4.6

Null S

subord

Table 2. Personalities’ results

3. DISCUSSION

In tables 1 and 2 the same kind of non-native constructions are present in both groups of

L2Ls for different types of production tasks. It is clear from these data that all of these

L2Ls have difficulties in identifying the subject in a native-like manner. They seem to be

entrenched non-native constructions, as confirmed by their perseverance in the speakers

with much higher exposure to robust L2 input and interaction, pointing to a well-

established underlying linguistic system whose specifications can hardly be modified and

its restructuring might only occur partially. On the other hand, comparing writing that has

been planned, which, arguably, might seem to be closer to some kind of learned

grammatical system of knowledge, with less planned spontaneous interview conversations

and still finding coincidences in the types of non-native constructions found may point to

a core of rules based on the interplay between the L1 and L2 underlying grammars and

discourse factors. Probably, the oral interviews reflect better the leaners’ attempts to

express their own meaning, leading more likely to fall back on their L1 as the conceptual

basis for articulation (Han 2010; Han & Lew 2012. (2

For both groups, there are fewer tokens of unanticipated inverted subjects with

unacussative, copula, passive or existential verbs than with it insertion (looking at the

average, half, approximately, for both groups. (2In some cases there are none, as for Bea

and Belle, which might be a sign of higher proficiency given the high number of interviews

sifted through for both (with Alda it is not so clear due to the smaller number of interviews

found. (2With complement or unergative verbs there are very few in either group: one

from student I of an embedded relative clause with a relativized complement: This shows

the capacity that *has the brain to process information; student X with two tokens of

quotative inversions with insist (e.g.:….insisted Yari); and students XI and XII with being

(e.g.: ...being their goal to affect the way in which they think), possible in their L1 but not

in the L2. There is also student IX with three examples of passives with there which, though

correct, sound very unnatural (marked with ?, like ?There have been selected repurposed

apps) as compared with their corresponding unanticipated ones with the passive subject

in canonical syntactic position.

Among the personalities, Jess, Gaby and Yoc present several cases of unanticipated

inversions with be and unaccusatives come and happen (e.g.: ...and came with the

congress the voting for impeachment; and then was one thing after another; happened

many crazy things); Raúl also with be, and Roz one token of an embedded relative clause

with a relativized complement and a transitive verb (I could see in him the effect that *was

producing my performance. (2Jess has the example This is something that made it possible

the human organization of humans where an expletive anticipates a NP object rather than

subject like all other cases encountered but the phenomenon is basically the same because,

again, only clausal objects can be anticipated this way. Thus, it seems that all participants

are aware of the thematic nature of the only (internal) argument of unaccusative

constructions keeping it verb internally but still having to fill the syntactic subject position

to the left of the verb as the L2 syntax calls for.

It is no wonder that the learners have difficulty with inversion for these verbs since the

absence of inversion in the L2 with unaccusatives and passives is less close to the

semantics than its presence in the L1. Underlying semantics point to the syntactic subject

as the internal (postverbal) argument of the predicate and placing it preverbally involves

its syntactic movement there. Besides, the learners might have received little negative

feedback to the contrary, especially the personalities as their interaction in the L2 has

probably been more focused on meaning than the students’. As in the case of null subjects,

the L2 does not provide evidence of the interdiction of postverbal subjects.

In their advanced EFL classes, the students have been made aware of and do

corresponding practice in, the use of anticipatory it only with clausal subjects and

unstressed there with NPs. Using these structures involves linguistic knowledge both at the

level of the syntax and its interface with discourse factors (focus) and phonetic ones (like

the end-weight principle of keeping more linguistic material after the verb. (2The fact that

these notional subjects may correspond to new information and so tend to appear

postverbally is independent of the provision of an anticipatory syntactic element which,

besides, is not necessary in their L1 (cf. Escutia 2008. (2The latter is rather a sign of

building an internal grammatical system with specifications of both the L2 (provision of

expletive it) and the L1 (inverted NP subject), as such an structure does not exist in the

L1 and the standard variety of the L2 -the one they work with in their classes and books-

requires a different expletive. Still, the IL of these students is moving within a possible

grammar since other (non-standard) varieties use the same expletive as our students in

all contexts (i.e. black vernacular English: it ain’t no heaven for you to go. cf. Labov 1969).

These types of non-native constructions are found both in the students’ planned writing

and in the older learner’s spontaneous or half-spontaneous oral production as some -few,

just for one speaker- of the data are also drawn from lectures, more amenable to following

a certain structure or an outline. Thus, a broad spectrum is covered of what has been called

the IL style continuum (Tarone 19839. (2This points to their being a stable part of their L2

competence as manifested in discourse processing and production along more (written) or

less (spoken) careful styles.

With respect to the production of null syntactic subjects either in main or subordinate

clauses, we find, as expected from their presence in many previous studies, that all our

L2Ls do produce them in discourse. The fact that we find fewer in the students is due to

the fewer data taken from them. Those of main clauses correspond mainly to postverbal

clausal subjects without expletive anticipation, especially in the case of the students, which

could be classified as just inversion of clausal subjects with copula verbs. There are also

some cases of null anaphora it whose reference can be recovered in the preceding

(students’) written or (personalities’) spoken discourse: e. g., several examples from

different personalities with depend (Maybe (

∅

) depends on the role I’m playing) and be (At

the beginning (

∅

) was very flattering for him. (2Those of subordinate clauses occur both

with adverbial and embedded clauses whose referents are in the main clause (e.g.: whichi

is really very good because (

∅

:iti) keeps you really humble; It represents the characteri in

the way that (

∅

: iti) should have been represented. (2In some cases they correspond to a

null anticipatory it of a postverbal clausal subject (e.g. I think (

∅

) is very nice once in a

while to talk to people who’s followed you all your life. (2As expected in cases of attrition,

there are inconsistencies within the same speaker: for example, Yoc and Raul produce,

respectively, the following utterances with internal inconsistencies in the presence of

referential it: No, because (

∅

) was my time, it was my life and … but it doesn’t depend on

me, (

∅

) depends on what they offer you10.

It may really be the case, then, that L2 syntax-discourse interface structures are more

amenable to implicit than explicit learning (as Han 2013 suggests) in the sense that

properties like inversion, which depend so much on discourse factors may not be taught

explicitly and can only be learned through robust input and interaction. However, they are

first regulated by the syntax of either language and are freer in word order in the learners’

L1. For example, with no context provided, many people came to the party can be

translated to the L1 either, word for word, as mucha gente vino a la fiesta or, with subject-

verb inversion, vino mucha gente a la fiesta, that is, with the subject placed pre or

postverbally. The latter is the proper answer in Spanish to the question what happened (at

the party)? with respect to the party, the stage topic, displaying focus on the postverbal

subject. The L2 might answer both that many people came to the party or, more formally,

rather in written discourse, that there came many people to the party. That is, in oral

discourse, the L2 displays inversion much less than the L1 because it is more syntactically

determined even in the kind of robust input that has been referred to, as it is quite formal

and literary. Thus, it seems that the L1 interface cues for inversion and keeping track of

subjects are overriding those of the L2, which are more syntactically determined. That

maybe partly why the interviews with the personalities do not seem to show much higher

proficiency in this area of inversion. Still, the fact that they tend to provide an expletive of

sorts to those inverted subjects shows that exposure to the L2 has left in all the learners

the awareness of the need for a preverbal syntactic subject.

The same could be said about the omission of pronominal subjects in the L1, which,

once their referent is mentioned in the preceding discourse, either in the main or

subordinate clause of a complex sentence, they can be omitted and identified by the verbal

inflectional morphemes. Thus, the learners may be using the discourse mechanism of

identification of their L1. At the syntactic level of the sentence they may not produce these

non-native constructions because, as advanced learners, they are proficient in the syntax

of the L2. It is those other interface aspects ocurring when one has to produce the language

in discourse that may be more influenced by the L1 discourse mechanisms of identification,

less amenable to explicit teaching and more permeable to L1 transfer.

As can be seen from the older learners’ speech, the it-insertion construction (iti + VP

(unacc./be/pass.) + NPi: take Alda’s It’s not important the appearance, the way you are

with your friends), also common in the planned written IL (and also in speech, though not

recorded here) of advanced adult Spanish learners of English in an institutional milieu

seems to persevere in the spontaneous oral production of older Romance learners with

long stays in English speaking countries and much more time of exposure and practice of

the language in natural contexts11. This points to its being a clear candidate for fossilization

in the English IL of this type of speakers. It seems to be more frequent than simple

inversion with the same type of verb -as it does not happen with other types. Their latter’s

occurrence might then point to a lower level of language proficiency as it does not

acknowledge the need of a necessary preverbal element. This may be consistent with the

fact that Belle and Bea, the apparently more proficient speakers, display no example of it

despite their many interviews consulted. This much lower frequency of unanticipated

inverted subjects might go against its clear status as a fossilizable type in favor of the

anticipated ones (*…if once happened in your life something like this (from Yoc) might

seem to be less proficient in the L2 than *it happened in your life something like this).

If, as it is likely, the input the EFL students in an instructed setting are exposed to is,

on the whole, non-robust, that is, infrequent and inconsistent, one may not rely on its

doing the job of correcting the L1 discourse biases about subject position and expression

with respect to new information and null subject production even though the syntax may

be in place. Still, Spanish students nowadays are exposed to a lot of English data, especially

these ones, who take an interest in being immersed in English because of their chosen

college major. Besides, many spend their summers in English speaking countries and watch

a lot of English videos. Furthermore, this type of students may develop a metalinguistic

consciousness about certain aspects of the language, in particular about the non-omission

of pronominal subjects as they are always warned about it. However, the strength of the

L1 discourse cues always at hand when one has to perform linguistically, along with both

the null semantic contribution of expletives and the ease of identification of coreferential

subordinate subjects may all cooperate in producing these non-native constructions.

Even though we have concentrated more on the it-insertion construction, some of our

learners retain both the (unanticipated) inversion of subject with unaccusative, copula and

passive verbs - though it seems that in smaller measure than the other- and the non-

native use of null subjects instead of expletive -and even referential- it in subordinate

subject coreferential clauses. All these seem to be manifestations of the same underlying

IL system, which legitimizes null (preverbal) syntactic subjects in discourse. It runs parallel

to how advanced English L2 learners of Spanish or Italian do not produce inversion when

it is appropriate -and even necessary- and overproduce subject pronouns in their Spanish

IL.

As explained above, the it insertion construction is consistent with the grammars of both

the L1, which prefers inversion with the types of verbs (unaccusatives and passives) and

corresponding discourse conditions, and the L2, with both possibilities of either placing the

NP subject before the verb or anticipating it with an expletive (there, in the standard

variety) and placing the NP post-verbally. Besides, it also occurs in the opposite direction,

that is, when the L1 is English and the L2 is Spanish: in this case the item chosen as

expletive may be the Spanish pronoun se and it looks like this structure tends to fossilize

as well, as explained above (cf. Escutia 2016. (2This IL provides an intermediate solution

whereby the same expletive occurs anticipating both NPs and clauses, as it occurs in other

natural languages like French (e. g. il existe la possibilité de les rencontrer /il parait qu’il

va faire de froid/il est mort le soleil ‘there exists the possibility of meeting them again/it

seems that it’s going to get cold/the sun has died’) or German (Es gibt nur Wasser überall/

Es ist interessant, dass niemand den Fehler bemerkt hat ‘there is only water everywhere/It

is interesting that nobody has noticed the mistake’); that is, using the default expletive it

in all cases, inserted in standard English only to anticipate clausal subjects12.

This it-insertion structure does not exist in the L1, so it is not originated in L1 transfer

but responds to a creative construction of the IL system, which resorts to it rather than

inversion or placing the subject NP preverbally when it is the focus of information. The fact

of its presence in the IL of advanced L2L’s of several L1 null subject romance backgrounds

may point to its stability as a fossilized idiosyncratic structure of null subject Romance IL.

These data seem to contradict those of Lozano & Mendikoetxea (2013) of acceptability

judgements of advanced (C2 proficiency level) EFL learners who accept it-insertion

structures at the same level as Ø-insertion, as compared with the corpora data they

analyzed, where overproduction of the former was rampant. Our own data conform more

with these corpora data, probably because when it comes to producing one’s own meaning

(thus, at the level of the interface syntax-discourse) L1 transfer seems to be more active.

For both sets of learners, we are moving within the confines of adult age with respect

to the beginning of significant exposure to the L2. The data might be quite different for

learners with as long but earlier exposure than our adult speakers, as some studies suggest

(cf. Johnson and Newport 1989, Long 1990, Newport 1991, Hartshorne,Tenenbaum, &

Pinker, 2018. (2We have to consider that our students have basically had an EFL instructed

exposure to the L2 from their early adolescence. They may also have probably listened to

a fair amount of Engish multimedially and only had some more intense face-to-face

interaction with native speakers in their vacation periods, which means that, if there really

is a sensitive period for L2A around adolescence, they were either past or in the middle of

it when they started with that impoverished exposure to the L2. No doubt the case of the

personalities is even clearer in this respect since two of them seemed to be in the same

situation as our students (Bea and Belle), and the rest really began their real exposure to

the L2 as adults.

It should have been very interesting to have presented both groups with an acceptability

judgement test of the items they had themselves produced or with equivalent ones as

more of the whole gamut of IL styles might have been covered. This would have given us

a more thorough picture of their knowledge about the type of items discussed. However,

as far as it goes, one can say that the personalities’ data may confirm the fossilizable status

of the types of items chosen (though less clearly so of the unanticipated inversion.

(2Another drawback is the lack of a comparison with what native speakers would do in

their interviews to see if one finds the same types of items.

We could have also added as candidates for fossilization the very common non-native

constructions of S-Aux inversion in embedded questions (Escutia 2002), Adjacency effects

with adverbials (cf. White 1990, Trahey and White 1993) and resumptive pronouns in

relative clauses (Gass 1979, 1982; Gass and Lee 2007), especially relativizing the subject

position (and also other positions), which might not have been expected from learners who

do not do so in their L1. All of these are also present in the interviews data of the

personalities. We have not taken them into account due to the fact that they were not so

frequent in our student’s written output (although we have observed them very often both

in their oral and written production) and because of the focus of our study on those types

of non-native constructions traditionally associated with null subject languages. We leave

them for possible future research.

Finally, when presenting non-native constructions such as these, one may feel the

responsibility of providing some remedial work to restructure the non-native linguistic

system somehow. Such work could be applied to these non-native constructions and doing

extensive in-class controlled practice with the students, making them aware of them as

used non-natively in real texts like the ones used here, apart from providing negative

evidence or correction when necessary. On the other hand, one may wonder if they are

the natural end for the type of learner studied here and no remedy is at hand except a

cosmetic one in those situations that naturally afford awareness and monitoring such as

written production or formal spoken tasks. Other speech situations do not lend themselves

to such awareness and control, which might anyway hinder naturalness and even

expressiveness in communication.

NOTAS

1 At the time they talked about (developmental) errors rather than non-native constructions and

also different from (performance) mistakes.

2 Unless it is filled with and adverbial of place or time: …and at that moment came the answer

from the other side; Near the house stood an huge pine tree.

3 We are referring here to, and our examples mainly deal with, what some authors consider to have

clausal or sentential focus, where one can postulate a “stage topic”, that is, an implicit topic which

signals the spatio-temporal parameters of the predication, the here and now of it (Erteschik‐Shir

1997).

4 These are the other type of discourse conditions dealt with in our examples, when a clause may

be pragmatically predicated of a topical referent not having at the same time a surface syntactic

topic, as illustrated in the comparison between the English dialogue and the corresponding Spanish

version:

P. What’s the problem with Michael? ‘¿Qué le pasa a Michael?

R. He’s just had an accident. ‘(*Él) acaba de tener un accidente

Here, in R. the syntactic subject (Michael), referred to by the pronoun he is the topic, in pragmatic

terms. In null-subject Romance languages, like Spanish, the syntactico-pragmatic conditions of the

language require for the third person pronoun él to be omitted seeming on the surface (as if there

is no apparent syntactic constituent as topic) even though there is clearly a topical referent signaled

by the verbal inflectional morphology.

5 At the C1 CEFR level, a language learner can: understand a wide range of long and demanding

texts or conversations; express ideas without too much searching; effectively use the language for

social, academic or professional situations; create well-structured and detailed texts on complex

topics.

6 In both groups, the corresponding tokens from this category correspond to a real pronominal

anticipation of a NP not mentioned before and not to an anaphoric element referring back to some

preceding referent which is repeated again for clarification (after a comma in writing and a small

pause in speech. (2

7 It is for this reason that no French L1 speaker has been chosen even though it is also a Romance

language and there are plenty of speakers of this kind that might have been picked. French is a

non-null subject language.

8 The same thing applies to the source and URL of the original interviews or lectures the data have

been drawn from. If a specialized researcher -one who is supposed to be discreet about it because

of their exclusive professional interest- is interested in the exact place and time where they can be

found, they can email the author of the article.

9 Tarone (1983) distinguishes different varieties within a speaker’s performance ranging from the

one produced for unattended speech data, which she calls the vernacular style, to a more careful

style when the speakers can monitor most their speech, with different varieties in between).

10 In the case of null subjects in general, we have tried not to use those with third person referential

or expletive it preceding the form is because in speech both might be conflated and difficult to

distinguish. Still, in a couple of cases that we have allowed them, the automatic subtitles of the

youtube screen have been activated to check for it trying to avoid that problem.

11 Although it only constitutes anecdotal evidence, I can witness to finding both myself, other

English teachers and advanced learners using at times such idiosyncratic construction in speech on

the spur of the moment.

12 It might be interesting to see if advanced L1 speakers of French and German do produce the it-

V-NP construction as well, which might reflect L1 transfer. If it is a question of transfer in this case,

our guess is that advanced speakers will be more conscious of it and will not do it.

REFERENCES

Al-Kasey, T. and Pérez-Leroux, A.-T., 1998. Second language acquisition of Spanish null subjects.

In Flynn, S., Matohardjono, G., O'Neil, W. (eds.), The Generative Study of Second Language

Acquisition. Hillsdale, N.J.: Lawrence Erlbaum, pp. 161-185.

Alexiadou, A., E. Anagnostopoulou & M. Everaert., 2003. The Unaccusativity Puzzle. Oxford

University Press.

Balcom, P., 1997. Why is this happened? Passive morphology and unaccusativity. Second Language

Research 13, 1, pp. 1–9. https://doi.org/10.1191/026765897670080531

Barbosa, P., 1995. Null Subjects. Ph.D. Dissertation, MIT. Cambridge, Mass.

Barbosa, P., 2000. Clitics: A Window into the Null Subject Property. In Costa, J., eds.), Portuguese

Syntax: Comparative Studies. New York: Oxford Press, pp. 31-93.

Belletti, A., Bennati, E. & Sorace, A., 2007. Theoretical and developmental issues in the syntax of

subjects: Evidence from near-native Italian. Natural Language Linguist Theory 25: 657–

689.

Burman, D., Bitan, T., Booth, J.R., 2008. Sex Differences in Neural Processing of Language Among

Children. Neuropsychologia: 46, 5, pp. 1349-1362.

https://doi.org/10.1016/j.neuropsychologia.2007.12.021

Domínguez, L. & Arche, M. J., 2014. Subject inversion in non-native Spanish. Lingua 145, pp. 243-

265.

Erteschik‐Shir, N., 1997. The Dynamics of Focus Structure. Cambridge: Cambridge University Press.

Ellis, R., 2003. Task-based language learning and teaching. Oxford University Press.

Escutia, M., 1998. Variability and non-variability in the English L2 acquisition of an adult Spanish

speaker. Perspectives on Spanish Linguistics, 2. Columbus, OH: Ohio State University, pp.

25– 43.

Escutia, M. 1999. Age and that-trace effects. Estudios ingleses de la Universidad Complutense 7,

pp. 109-118.

Escutia, M., 2002. Universal Grammar, transfer and variability: a case study. Estudios ingleses de

la Universidad Complutense 10, 2002, pp. 67-86.

Escutia, M., 2002. First language effect on Spanish learners’ production of indirect interrogative

clauses. Círculo 10. 1-5. Available at http://ww.ucm.es/info/circulo/no10/index.htm

Escutia, M., 2008. Transfer and Universal Grammar in Unaccusative Constructions Errors.

Miscelánea, 37, pp. 23-38.

Escutia, M., 2010. El uso de "se" con verbos inacusativos por estudiantes avanzados de español

como lengua extranjera: transferencia y restructuración. Revista Española de Lingüística

Aplicada 23, pp. 129-152

Escutia, M., 2013. Expletives and unaccusative predicates in L2A. Higher Education of Social science

2, pp. 1-14.

Escutia, M., 2016. L2 Spanish preverbal ‘se’ in analysis and production data. Revista Española de

Lingüística Aplicada, pp. 30-63

Fernández Soriano, O., 1989. Strong pronouns in null-subject languages and the avoid pronoun

principle. MIT working papers in linguistics 11, pp. 228–239.

Filiaci, F., Sorace A. & Carreiras, M., 2014. Anaphoric biases of null and overt subjects in Italian and

Spanish: A cross-linguistic comparison. Language, Cognition and Neuroscience 29(7), pp.

825–843.

Gass, S., 1979. Language transfer and universal grammatical relations. Language Learning 29, pp.

327–344.

Gass, S., 1982. From theory to practice. In Hines, M. & Rutherford, W., eds.) On TESOL '81: 129–

139. Washington DC: TESOL.

Gass, S., 1997. Input, Interaction, and the Second Language Learner. Mahwah, NJ: Lawrence

Erlbaum Associates, Publishers

Gass S. and Lee, J., 2007. Second Language Acquisition of Relative Clauses. Studies in Second

Language Acquisition. Volume 29, 2, pp. 329-335

Han, Z. -H., 2004. Fossilization in Adult Second Language Acquisition. Clevedon, England:

Multilingual Matters.

Han, Z.-H., 2006. Fossilization: Can grammaticality judgment be a reliable source of evidence?. In

Han Z.-H. & Odlin, T., eds.) Studies of fossilization in second language acquisition, pp. 56–

82. Clevedon, England: Multilingual Matters.

Han, Z.-H., 2009. Interlanguage and fossilization: Towards an analytic model. In Cook, V. & Wei,

L., eds.) Contemporary applied linguistics I: Language teaching and learning. London:

Continuum, pp. 137-162.

Han, Z.-H., 2013. Forty years later: Updating the Fossilization Hypothesis. Language Teaching 2,

pp. 133 -171. https://doi.org/10.1017/S0261444812000511.

Han, Z.-H. & Lew, W.M., 2012. Acquisitional complexity: What defies complete acquisition in SLA.

In Szmrecsanyi, B., & Kortmann, B., eds.), Linguistic Complexity in Interlanguage Varieties,

L2 Varieties, and Contact Languages. Berlin/New York: Walter de Gruyter, pp. 192-217.

Hartshorne, J. K., Tenenbaum, J. B., & Pinker, S., 2018. A critical period for second language

acquisition: Evidence from 2/3 million English speakers. Cognition, 177, pp. 263– 277.

https://doi.org/10.1016/j.cognition.2018.04.007

Hertel, T.J., 2003. “Lexical and discourse factors in the second language acquisition of Spanish

word order. Second Language Research 19, pp. 273- 304.

Hilles, S., 1986. Interlanguage and the pro-drop parameter. Second Language Research 2, pp. 33-

52.

Johnson J.S. and Newport E.L., 1989. Critical period effects in second language learning: the

influence of maturational state on the acquisition of English as a second language. Cognitive

Psychology 21 (1), pp. 60-99.

Kato, Mary A., 1999. Strong and weak pronominals and the null subject parameter. Probus 11(1.

pp. 1–38.

Labov, W., 1969. The Logic of Non-Standard English. Georgetown Monographs on Language and

Linguistics, 22, pp. 1-31.

Lardiere, D., 1998. Dissociating syntax from morphology in a divergent L2 end state grammar.

Second Language Research, 14 (4), pp. 359–375.

Lardiere, D., 2007. Ultimate Attainment in Second Language Acquisition: A Case Study. New York:

Routledge

Liceras, J. M., 1988. Syntax and stylistics: more on the pro-drop parameter. In Pankhurst, J.

Studies on Language Acquisition, 7. De Gruyter Mouton

Liceras, J. M., 1989. On some properties of the pro-drop parameter: Looking for missing subjects

in non-native Spanish. In Gass, S. & Schachter, J., eds.), Language accquisition: A linguistic

approach. Cambridge: Cambridge University Press, pp. 109–133.

Liceras, J. M., & Díaz, L., 1999. Topic-drop versus pro-drop: null subjects and pronominal subjects

in the Spanish L2 of Chinese, English, French, German and Japanese speakers. Second

Language Research, 15(1), pp. 1–40. https://doi.org/10.1191/026765899678128123

Lightbown, P. M., 1985. Great Expectations: Second-Language Acquisition Research and Classroom

Teaching. Applied Linguistics, 6, 2, pp. 173–189. https://doi.org/10.1093/applin/6.2.173

Lightbown, P. M., 2000. Anniversary article: Classroom SLA research and second language teaching.

Applied Linguistics 21, pp. 431-462.

Lightbown, P. M., 2003. SLA research in the classroom/SLA research for the classroom. Language

Learning Journal 27, pp. 4-14.

Long, M. 1990. Maturational constraints on language development. Studies in second language

acquisition 12, pp. 251–85.

Long, M. H., 2003. Stabilization and fossilization in interlanguage development. In Doughty, C. J.

& Long, M. H., eds.), Handbook of Second Language Acquisition Malden, MA: Blackwell, pp.

487–535.

Lozano, C., 2003. Universal Grammar and focus constraints: The acquisition of pronouns and word

order in non-native Spanish. University of Essex: Unpublished PhD dissertation.

Lozano, C., 2006a. “Focus and split intransitivity: The acquisition of word order alternations in

non-native Spanish. Second Language Research 22(2): 145–187.

https://doi.org/10.1191/0267658306sr264oa.

Lozano, C., 2006b. The development of the syntax-discourse interface: Greek learners of Spanish.

In: Torrens, V. & Escobar, L., eds.), The acquisition of syntax in Romance languages.

Amsterdam: John Benjamins, pp. 371-399.

Lozano, C., 2009. Selective deficits at the syntax-discourse interface: Evidence from the CEDEL2

corpus. In Leung, Y., Snape, N. & Sharwood-Smith, M., eds) Representational Deficits in

Second Language Acquisition. Amsterdam: John Benjamins, pp. 127–166.

Lozano, C. & Mendikoetxea, A., 2013. Corpus and experimental data: Subjects in second language

research. In Granger, S., Gilquin G. & Meunier, F., eds), Twenty Years of Learner Corpus

Research: Looking back, Moving ahead. Corpora and Language in Use. Proceedings 1,

Louvain-la-Neuve: Presses universitaires de Louvain, pp. 313-323.

Mendikoetxea, A. & Lozano, C., 2018. From Corpora to Experiments: Methodological Triangulation

in the Study of Word Order at the Interfaces in Adult Late Bilinguals (L2 learners. Journal of

Psycholinguistic Research 47(02), pp. 411-430.

Montrul, S. and Bowles, M., 2009. Back to basics: Differential Object Marking under incomplete

acquisition in Spanish heritage speakers. Bilingualism: Language and Cognition, 12, 3, pp.

363–383.

Montrul, S. & Rodríguez-Louro, C., 2006. Beyond the syntax of the null subject parameter: A look

at the discourse-pragmatics distribution of null and overt subjects by L2 learners of Spanish.

In Torrens, V. & Escobar, L., eds.), The acquisition of syntax in Romance languages.

Amsterdam: John Benjamins, pp. 401–418.

Moskovsky C. and Ratcheva, S., 2014. L2 Fossilization: A Competence or a Performance

Phenomenon. The Open Communication Journal 8, 9-17.

Newport, E., 1990. Maturational constraints on language learning. Cognitive Science 14, pp. 11–

28.

Oshita, H., 2001. The unaccusative trap in second language acquisition. Studies in Second

Language Acquisition 23(02), pp. 279 – 304. https://doi.org/10.1017/S0272263101002078.

Pérez-Leroux, A., Glass, W. L., 1999. Null anaphora in Spanish second language acquisition:

probabilistic versus generative approaches. Second Language Research, 15(2), pp. 220–

249. https://doi.org/10.1191/026765899676722648

Phinney, M., 1987. The pro-drop parameter in second language acquisition. In Roeper, T. and

Williams, E., eds), Parameter setting. Dordrecht: Reidel: pp. 221-38.

Pladevall, E., 2013. Adult L2 Spanish development of syntactic and discourse subject properties in

an instructional setting. RAEL (Revista Electrónica de Lingüística Aplicada) 12, pp. 111-129.

Saville-Troike M., 2006. Introducing Second Language Acquisition. Cambridge Introductions to

Language and Linguistics.

Selinker, L., 1972. Interlanguage. International Review of Applied Linguistics, 10, pp. 209–231.

Singleton, D., 1989. Language acquisition: the age factor. Clevedon: Multilingual Matters.

Singleton, D., 1995. Introduction: a critical look at the critical period hypothesis in second language

acquisition research. In Singleton, D. and Lengyel, Z., eds), The age factor in second language

acquisition, Clevedon: Multilingual Matters: pp. 1–29.

Sorace, A., 2000. Syntactic optionality in non-native grammars. Second Language Research, 16

(2), pp. 93–102.

Sorace, A., 2003. Near-nativeness. In C. J. Doughty & M. H. Long (eds.), Handbook of Second

Language Acquisition (pp. 130–151. Malden, MA: Blackwell.

Sorace, A., 2004. Native Language Attrition and Developmental Instability at the Syntax-Discourse

Interface: Data, Interpretations and Methods. Bilingualism: Language and Cognition 7(02),

pp. 143 – 145. https://doi.org/10.1017/S1366728904001543

Sorace, A., 2005. Selective optionality in language development in Syntax and Variation:

Reconciling the Biological and the Social. In Leonie Cornips and Karen P. Corrigan (eds.)

Current Issues in Linguistic Theory 265, pp. 55–80.

Sorace, A. & Serratrice, L., 2009. Internal and external interfaces in bilingual language

development: Beyond structural overlap. International Journal of Bilingualism 13(2.

https://doi.org/10.1177/1367006909339810

Tarone, E., 1983. On the Variability of Interlanguage Systems. Applied Linguistics, 4, 2, pp. 142–

164. https://doi.org/10.1093/applin/4.2.142

Tsimpli, I., Sorace, A., Heycock, C. & Filiaci, F., 2004. First language attrition and syntactic subjects:

A study of Greek and Italian near-native speakers of English. International Journal of

Bilingualism, 8, 3, pp. 257-277.

https://doi.org/10.1177/13670069040080030601

Trahey, M. and L. White., 1993. Positive evidence and preemption in the second language

classroom. Studies in Second Language Acquisition 15, 181–204.

Trenkic, D., 2009. Accounting for patterns of article omissions and substitutions in second language

production. In M. P. García Mayo & R. Hawkins (eds.), Second language acquisition of

articles: Empirical findings and theoretical Implications. Amsterdam: John Benjamins, pp.

115–143.

White, L., 1985. The pro-drop parameter in adult second language acquisition. Language Learning

35, pp. 47-62.

White, L., 1990. The verb-movement parameter in second language acquisition. Language

Acquisition 1, pp. 337–60. https://doi.org/10.1207/s15327817la0104_2

White, L., 2003. Second Language Acquisition and Universal Grammar. Cambridge: Cambridge

University Press.

Zobl, H., 1980. The formal and developmental selectivity of L1 influence on L2 acquisition.

Language Learning, 30, 43-57. https://doi.org/10.1111/j.1467-1770. 1980.tb00150.x

APPENDIX I: STUDENTS’ WRITTEN DATA

LEARNER

INVERSION no

expletive

INVERSION it-

insertion

NULL SUBJECT Main

clause

NULL S Subordinate

Student I

First there is cooing,

then starts babbling

After this stage

appears the one-word

stage

In this stage come

into play mistakes

This shows the

capacity that has the

brain to process

information

Then arises the

question of how much

can language dictate

what we think

Later it arose the idea

that Eskimos have

certain prefixes

In English and

Indonesian it is used

the expression a long

time

Then it is found a

summary of the

arguments

It must be borne in

mind the following

details: …

It is noticeable the

wrong use of the

particle

The more

transformations the

more difficult (∅) is

for us to process the

sentence

In English (∅) seems

to be natural to have

the pronouns he,

she, it

it is dangerous in the

sense that (∅) might

provoke undesirable

responses

Student

They have exposed

the idea that exist

certain central

metaphors

In the Work of Tannen

it has been done a

differentiation

between the two

types of

conversational styles

It has also been

researched the

tendency that

subordinates have to

use indirect speech

Then it arises the

question of whether

or not our mind is

shaped by language

… just because you

speak differently (∅)

doesn't mean you

think different

Student

III

Thought is possible

without language,

being babies an

example of this

Thought and

language are

different, being

language the means

by which we

communicate

thoughts

It will be explained

the possible

connection between

language and thought

It will also be

mentioned the

different ways in

which languages

express different

concepts …..

There is a part in

which it is described

the variety of terms

used.

In Lackoff and

Johnson (∅) is

argued that there

are central

metaphors

Unlike English, Spanish

only speaks of

intentional actions

when (∅) is obvious

that someone did it

Student

In Russian are used

different adjective

endings

It has been

researched the subtle

shades of meaning

that different word

orders convey

It's interesting the

thing that speakers

living in Miami

probably have a

minimal vocabulary

for snow

Thus, it is interesting

the way in which

indirect speech can

be used

It was made an

experiment which

consisted in teaching

speakers a new way

to talk about time

In that simple

sentence (∅) is

needed to mark the

masculinity of the

chair many times

after reading this

info (∅) is clear that

speakers of different

languages think

differently

For instance, in

English (∅) is needed

to change the verb to

know if the action

happened in the past

… and one has to

mark if (∅) is

something that the

person has heard

To discuss the belief

that (∅) is not language

what shapes our

thoughts

You need to be more

specific and mention

what kind of tree (∅) is

They showed images in

which (∅) was clear

who made the action

Student

Not only was

described the

theoretical

background there

There was a

difference between

oral language and

reading

comprehension, being

this identified as the

problem

It is avoided possible

confusions in the

terminology

It has been opted for

the one suggested by

E Anthony

However, it will be

tackled the methods

which had an impact

It is given priority to

the teaching of

speaking

It was created a

language learning

program

It is only used the

target language

it is taken into

account the

unconscious elements

involved

It is needed an

atmosphere in which

learners feel

comfortable

It is preferred fluency

to accuracy

It is suggested the

employment of songs

To understand

emotional meaning

(∅) is necessary to

comprehend the

vocabulary

The principal

advantage of this

approach is that (∅)

allows the possibility of

including other

methods

… since the learners

decide the materials

that (∅) will use in the

classroom

It is recommended

the utilization of body

answers

Student

… to learn a foreign

language, being

English the language

in this case

It was begun to take

into consideration the

cultural factors

It is presented/

defended the idea

that language must

be treated as a

whole/ that each

person has 8 different

intelligences

It has been taken into

account any kind of

factor that could

influence language

learning

There is no method

that could be

considered the best

and probably it will

never be

It existed a need for

new approaches

in German the Sun is

feminine and in Spanish

(∅) is masculine

Student

VII

For the learning of a

foreign language are

necessary two

processes

This is the way how

appears the study of

other methods

A subject in which is

taught the different

phenomena of the

language

It happened

something similar

It will be included

information about the

Orton-Gillingham

approach

if it is simultaneously

used the multiple

sensory methods

It should be avoided

long texts

It exists the

conception that

dyslexia can be cured

In sum, children even

though (∅) repeat to

some extent words and

phrases create

language ….

Taking into account the

whole learning that

implies any word and

the relation to each

component that (∅)

represents, we can

say…

Student

VIII

In the next point are

going to be developed

some of the activities

that use multisensory

techniques

With dyslexic children

might be used

activities that use

some of their senses

It is also found the

previously mentioned

taboo words

It occurred bigger

changes in word

formation

It occurred bigger

changes in word

formation

There are advanced

students who really feel

that (∅) are bad at

breaking words

In this point are

presented some

general exercices

You can create a box

in which appear words

with these sounds

...using fun

worksheets in which

appears a legend

with simple

instructions

It occurred the

process of

specialization

It is found such words

as hobnob

It has been explained

the different types of

semantic change

Student

In the year 731 in

which can be found

historical events such

as the historical work

of Beda

Sometimes it may

occur the opposite

process

There have been

chosen popular

applications oriented

to language learning

There have been

selected repurposed

apps

There have been

created five groups

To understand the

history (∅) is crucial

not only to know….

Student

Visually could be

observed rare ocular

movements

In a state of

emotional discomfort,

insisted Conture

On the one hand,

appears a gradual

start of stuttering

Tension while

producing language,

insisted Yairi

It can also occur other

types of interruptions

In this paper it will be

used the term person

who stutters

It has been found an

excessive tension

while stuttering

It was analyzed

different strategies

that patients used

It could also be found

more phonemes

where the subject

stutters

In this test it can be

seen three different

punctuations given

Now that (∅) is clear

what kind of

symptoms can

appear, let’s

continue with …

…but for a person

who stutters (∅)

might be a real

struggle to form the

sound

The technique is called

semantic feature

analysis for the reason

that (∅) is a vocabulary

strategy to help

students

Student

...being their goal to

affect the way in

which they think

Now it will be

analyzed the

conversation

It can be clearly seen

the influence of their

L1.

When the air controller

listens to the

communication, (∅)

hesitates about the

pilot’s indecision

…but it could have

been a slight decrease

at the end of the unit

Student

XII

…being the best

education in nursery

school

In the last last day will

be the awards

ceremony

it can also be found a

chapel in that school

It was also really

noticeable the

separation in class

among the different

“tribes”

it will be stated a

hypothesis on the

possible final results

First of all (∅) is

important to choose

a topic of their

interest

Tell the students that

(∅) is a good option to

do it about Christmas

APPENDIX II: PERSONALITIES’ DATA

LEARNER

INVERSION no

expletive

INVERSION it-

insertion

NULL S MAIN

CLAUSE

SUBORDINATE NULL

YOC

(L1

Peninsular

Spanish)

But if once

happened in your

life something like

this (2009)

Happened many

crazy things in my

life (2010)

It (=there) will be

some time in the

future for us to be

together (1999)

So, when it came

that situation and

starting to have

time to play (2012)

I can play for all

kinds of ages: It’s

incredible thát

(2014)

It was great the

opera of Dubai

(2018)

For me, (∅) was not

an easy decision

because I had

already three kids

(2010)

So, (∅) was

something that I

didn’t think (2012)

I think so, (∅) was a

very different

situation (2014)

No, because (∅) was

my time, it was my

life (2014)

The girls, (∅) is something

that (∅: they, people)

think is very interesting

(2014)

I adore him because (∅)

was an incredible singer

(2015)

RAÚL

(L1

Peninsular

Spanish)

When you put two

Spaniards together

usually are long

nights

(2010)

And then are those

characters that I

want to put more

attention to (2012)

It has to happen

something that is

out of control

(2007)

It’s spectacular the

rugby nowadays

(2008)

It’s like this the

world (2008)

And that’s why it’s

important friends

and family because

they remind you

who you are (2012)

Because somebody

will see it and (∅) will

be meaningful to that

person (2010)

But it doesn’t depend

on me, (∅) depends

on what they offer

you (2012)

For me, (∅) was one

of those moments

where I was pinching

myself (2017)

The other day (∅: I)

was talking to Asgar

about it (2018)

We’re working in this

movie that I think (∅) is

gonna be a good one

(2012)

I wanted to know where

(∅: it, the character) was

coming from (2015)

I think that (∅: it, his

attitude) has to do with

my mom (2017)

…in other words, that (∅:

it) is based on the

material.

(2017)

I think it’s

important the body

language (2012)

They just are with all

the consequences.

And usually (∅: they)

are people who are

very loving and

caring for others

(2018)

For me (∅) was a

great moment

because I really

admire his work

(2018)

If you want to see the

world of narcotrafic as (∅:

it) was invented (2017)

I knew that (∅: it) was

never going to happen

(2018)

Because, again, (∅)

shows you that the

strongest, the people with

real talent are good

people, usually (∅) are

great people endowed

with empathy (2018)

JESS

(L1

Peninsular

Spanish)

Organisms which

can live only by

killing other

organisms, which

(∅: there) have

been several of

these cases (2012)

If I run the program

again (∅: there)

come different

outcomes (2013)

In some respects

more important are

behavioural

differences

between us and the

apes (2014)

In the final decade

(∅: it) had been

discovered that

where the land is

exposed but if there

has been this kind

(2015)

...that made it

possible the human

organization of

humans (2014)

In the old times (∅:

it) was explained as

the result of the

intentional action of a

Supreme Being

(2012)

Sometimes in the

case of religion (∅:

it) has to do with the

interpretation (2013)

For all the examples

he gave, (∅: it) has

been shown how the

evolution occurred

(2013)

Moreover (∅: it,

molecular biology)

allows us to

reconstruct the

history of living

organisms (2015)

For much of time (∅:

it, the brain) was like

that, about 600

grams (2016)

But I think that by

and large (∅: there)

will be progress

(2016)

Wilmut pointed out that

(∅: it) had taken him 270

efforts to clone Dolly

(2012)

So the best they can do

for the genes is to be

eaten by the female so

that (∅: she) is very

healthy, produces more

healthier babies (2012)

The theory of intelligent

design accomplishes

exactly the opposite of

what (∅: it) intends

(2013)

What bipedalism implies

is that (∅: it) leaves the

arms free (2014)

Those genes will be

favored because (∅) are

more likely to be

transmitted to more

progeny (2014)

The only animal that is

aware that (∅) exists as

an individual (2014)

Those things will be

favored because (∅:they)

are more likely to be

transmitted to more

progeny (2014)

…so (∅: he, Hume)

derives moral principles

from facts (2014)

Gorillas and chimpanzees

cannot speak, not only

because (∅: they) don’t

have the ability but also

because they cannot emit

the sounds we can (2016)

…if (∅: it, life) happened

long ago then I mean a

few thousand years

(2017)

If (∅) had occurred we

would know about it

(2017)

A person living in

Scandinavia can’t do very

well because (∅: she)

cannot synthesize vitamin

D (2017)

GABY

(Brazilian)

that’s when came in

the bottle the

message that he

needed to talk to

me (2016)

And came with the

congress the voting

for impeachment

(2016)

And then was one

thing after another

(2016)

And then was the

same director who

invited me to do

acting (2016)

And then became

part of my life being

recognized in the

street (2016)

What’s wrong in

Brasil is that it

should be a

dialogue and not

like a country

divided (2016)

It’s happening

something right

now: that Brazil is

divided (2016)

The present, which

is now the past

because it passed a

year (2017)

When you’re not

representing a

character as a star

that’s when (∅)

should matter (2016)

And when (∅: what

you have inside)

goes to the

audiences, it’s theirs

too

(2016)

…and that’s why (∅:

the film) becomes so

strong (2017)

Why did you impose that

(∅: it) would be me to do

that character (2016)

He showed me where (∅:

it) was gonna be (2016)

I think that (∅: it) came in

the time when we needed

something and that film

represents that

resistance… and as long as

(∅: there) is democracy

we have the right (2016)

… because (∅: it, the film)

became in a way

connected with that story

(2016)

…. because (∅: it, the film)

touched them in that way

(2016)

I think that (∅: it, the film)

brought back columns of

discussion (2016)

Why Sao Paulo was so

ugly because (∅) was so

big and everything…

(2016)

And there is a long silence,

so (∅) creates some

feeling that you don’t

know what’s gonna

happen (2016)

He told me (∅: it) was

gonna be relaxed and it

really was (2016)

The other day that was the

occupy movement that

they took over the city, so

(∅: it) demonstrated that

the city was not the few

makers in the art (2016)

That’s when I think that

(∅: her call to be an

actress) started (2017)

I think that (∅: it) is the

audiences that decides

(2016)

I think (∅: there) would be

movies by then (2016)

Because I thought (∅: it)

could be one of the friends

(2017)

BEA

(Mexican

Spanish)

…because it’s being

stolen all the money

(2006)

It’s really not

investigated the

wonders of the

other half (2015)

It must also be

exciting for men

what can be our

contribution (2015)

It did not help that

part (= that part

didn’t help) (2015)

It was divided the

school between the

good nuns and the

bad ones (2017)

It was too clear your

thought process

(2017)

It makes no sense

the character

(2017)

It doesn’t work

because it’s still

very strong his

accent (2017)

…but for the youth

(∅: it, this fact) is

very important

(2015)

Because my father was

very jealous, (∅) didn’t

allow me to go to college

until I turned 16 (2015)

I think that (∅) is

important to tell our

stories (2017)

It just keeps feeling

strange this movie

(2017)

ROZ

(L1

Peninsular

Spanish)

I could see in him

the effect that was

producing my

performance

(2019)

It was too big the

light that the

character produced

(2018)

And then it came to

me what he was

explaining before

(2019)

The happiest time for

me in America was

not in front of a

camera, (∅) was in

49th street (2015)

In a way (∅) has to

do with my own story

(2017)

For us (∅) was very

important to have an

institution behind

At the beginning (∅)

was very flattering

for him (2018)

For me has been very

important all the

work that I have

done with him

(2019)

For me was very

important all the

silences of the

character (2019)

The meaning of the bull is

that (∅) is a bull (2018)

… and the other issue I

thought that (∅) was faith

(2015)

I am very grateful to my

profession not for what (∅:

it) gave me in terms of

being famous…but in

terms of what it taught me

as a human being (2015)

It represents the

character in the way that

(∅: it) should have been

represented (2018)

I like to be precise

because (∅: there) may be

movies that have not been

very strong that year but

they have a fantastic

soundtrack (2018)

For me is almost

impossible to verbalize

how emotional (∅: it) is to

have reached this point

with this movie (2019)

It was one of the best

experiences in my life

because (∅) actually

established a new order

(2019)

BELLE

(L1

Peninsular

Spanish)

It’s really inspiring

everything that he

gives you (2012)

It’s very intriguing

this picture (2018)

(this time) (∅: it) was

a different man but

the same behavior

(2012)

…but (∅: it) depends,

in every character I

use different things

to prepare (2018)

Yes, in a way (∅: it)

can be a comfort

zone but in a way, it

is more delicate

(2018)

… I like that (∅: it) is not

a formula, that is not

really related to that

(2013)

Which is really very good

because (∅) keeps you

really humble (2018)

Because that day I think

(∅: he) had a semi-nude

scene (2018)

And I think that (∅) is as

important for him as the

first day (2019)

ALDA

(Italian)

It has spread so

much the

Mediterranean diet

Maybe (∅) depends

on the role I’m

playing (1977)

You cannot imitate

Chaplin because (∅) is

unimitable (1977)

that it is wonderful

(2012)

It (: there)’s

something

incredible that it’s

happening to me

(2016)

It’s not important

the appearance;

the way you are

with your friends

that’s life, the good

life (2020)

Sometimes was

difficult to work with

him (2014)

…and it’s now that I feel

that (∅) is gone forever

(2009)

I think (∅) is very nice

once in a while to talk to

people who’s followed you

all your life (2016)

It’s so moving sometimes

that (∅) really brings up

tears (2016)

MADDY

(Italian)

…and so it was

really a beautiful

experience this

movie (2013)

I don’t know what it

means really to

have a base and I

don’t know how it’s

gonna be my life

next year (2013)

… and also Keanu

Reeves, so (∅) was a

very big cast (2013)

If it’s a big budget or

not (∅) doesn’t really

matter (2013)

I love animals but,

because of that, (∅)

was easy to learn

(2016)

Actually (∅) was

funny at the

beginning (2016)

For me, (∅) was like

the possibility to

back in film (2018)

I started out as a model

because (∅) was the

easiest way to escape

(2010)

…because Italian movies

are not anymore like (∅)

used to be (2012)

When I work with him I

don’t think that (∅) is my

husband (2012)

I think more of the

process of creation

because was really, it was

really the process of

creation (2013)

When we did the rape

scene was actually easier

that way (2010)

… and was really

interesting to work with

Sophie and really there

was a beautiful chemistry

between us and was

beautiful to go to Cannes

together (2016)