CONJUGATION OF IRREGULAR VERBS BY

NATIVE BRAZILIAN PORTUGUESE SPEAKING

CHILDREN: A STUDY IN THE LIGHT OF LEXICAL
PHONOLOGY AND MORPHOLOGY

Tamires Pereira Duarte GOULART *

Carmen Lúcia Barreto MATZENAUER **

■■ ABSTRACT: This study investigates the acquisition of irregular verbs in Brazilian Portuguese
(BP) when conjugated in the Indicative Present, Subjunctive Present and Simple Past tenses
by native BP speaking children. It aims at describing and analyzing the interaction between
Morphology and Phonology which is found in the inflection of irregular verbs, in the light
of Lexical Phonology and Morphology (LPM). Data show that linguistic phenomena of
morphophonological origin, which are highly complex, can be considered processes of late
acquisition by Brazilian children. LPM leads to the understanding that verb regularization, one
of the manifestations of irregular conjugations, can be explained by the non-correspondence
between the phonological and morphological levels of the language.

■■ KEYWORDS: morphophonological acquisition; irregular verbs; Lexical Phonology and

Morphology

Preliminary Considerations

In the process of language acquisition, from the input they receive, children must
abstract the grammar of the language, that is, the units that constitute each component of
the linguistic system, such as phonology, morphology and syntax, as well as the relations
among them. Relations can occur either between units within the same component
or between units of different components. The latter implies interfaces between the
components that make up the grammar of the language.
Considering relations between the components of the grammar, this study focuses
on the interface that is established between Phonology and Morphology and aims at
the behavior, in the phase of language acquisition, of the adjunction of inflectional

*
Catholic University of Pelotas (UCPEL), Postgraduate Program of Letters, Pelotas – Rio Grande do Sul – Brazil.
tamirespdgoulart@gmail.com
**
Catholic University of Pelotas (UCPEL), Postgraduate Program of Letters, Pelotas – Rio Grande do Sul – Brazil.
carmen.matzenauer@gmail.com

Alfa, São Paulo, v.62, n.1, p.169-189, 2018 http://doi.org/10.1590/1981-5794-1804-8 169

This work is licensed under a Creative Commons Attribution 4.0 International License.
morphemes to irregular verb forms of Brazilian Portuguese (BP) in cases with
alternations in the phonological plane of the language.
The aim of this paper is to discuss the use of forms of the Indicative Present,
Simple Past and Subjunctive Present, which, in verbs classified as ‘irregular’, show
alternation of consonant phonemes. Examples of such verbs are fazer, trazer and medir
(‘to do’, ‘to bring’ and ‘to measure’), which, in the previously mentioned tenses, show
the following alternations: [s] alternates with [z], as in faço, fazes (fa[s]o, fa[z]es);
[g] alternates with [z] and [s], as in trago, trazes, trouxe (tra[g]o, tra[z]es, trou[s]e);
[s] alternates with [d], as in meço, medes (me[s]o, me[d]es).
Research data, which was collected by an instrument that elicited the use of forms
in the verb tenses under study, were analyzed and formalized with the support of Lexical
Phonology and Morphology, since it is a theory that predicts the interaction between
Morphology and Phonology, pointing out relations between the morphological structure
of a word and the phonological rules that apply to it.

Irregular verbs and consonantal phonological alternations

The morphological component of Brazilian Portuguese allows regular and irregular

verbs. For their characterization, it should initially be pointed out, following Câmara
Jr. (1970), that a verbal is composed of a theme (T), formed by a root (R) and a theme
vowel (TV), besides inflectional suffixes (FS), which can be of mood and tense (SMT)
and number and person (SNP). Considering this structure, verbs are said to be regular
when they keep an unchanged root and follow the general pattern in the adjunction of
inflectional morphemes; it is the case of the verb amar (‘to love’) in all its conjugations:
amo, amava, amarei, amasse.
Differently, irregular verbs, the focus of this paper, are those in which there is a
change in the root when they are conjugated. Examples are the verb dizer (‘to say’),
and its inflections digo and dizes (di[g]o, di[z]es), and verbs that deviate from the
paradigm imposed by the grammar of the language, thus, showing changes in inflectional
morphemes. The verb saber (to know), whose conjugation in the Indicative Present
has the forms sei, sabes, sabe, is an example of this type of irregularity. The verb form
“sei” moves away from the paradigm of the first person of the singular, conjugated in
the Indicative Present, that usually manifests with the personal-number suffix -o, as in
canto, danço, escrevo, faço, estudo.
In irregular verbs, either due to changes in the root or because of escape to the
paradigm, alternations may occur. In the functioning of the language, the phenomenon
of alternation is characterized by replacing a phoneme by another and there are two
types of alternations: vowel alternation, which implies the exchange of vowels, for
example, [o]vo ~ [ɔ]vos, and the consonantal alternation, which is understood as the
replacement of a consonantal phoneme with another one. Examples are consonantal
changes at the end of roots of conjugated verbs, as in fazer ~ faço ~ fazes (fa[z]er, fa[s]
o, fa[z]es); trazer ~ trouxe ~ traz (tra[z]er, trou[s]e, tra[s]).

170 Alfa, São Paulo, v.62, n.1, p.169-189, 2018

 The previously described consonantal alternation seems to be highly complex for
BP speakers. As a result, it is argued that its acquisition can be considered a process
of late nature in the acquisition of the BP irregular verb system of BP that extends, at
least, to the age of 10.

The process of regularization of irregular verbs

The tendency towards regularization and at less marked in the functioning

of the grammar of languages is recognized as part of the evolutionary process of
linguistic systems, as well as of the process of language acquisition. In this context,
the regularization of verb forms is inserted.
Therefore, there is regularization of irregular verbs when the speaker produces
an irregular verb in agreement with the pattern of the regular verb. Examples are the
production of “fazi” instead of “fiz” (‘did’) and “trazi ~ trazei” instead of “trouxe”
(‘brought’).

Lexical Phonology and Morphology (LPM)

Lexical Phonology and Morphology (LPM) is a theory that enables researchers to

look at language while considering the relationship between morphology and phonology.
As a result, generalizations and patterns of morphophonological occurrences can be
singularly captured.
Linguistic systems show cases of interface among the components of the language.
So, neither in the phonological scope nor in the morphological one alone, would it be
possible to explain them. An example, in this context, is the phenomenon under study,
which puts into analysis the production of irregular verbs that exhibit phonological
alternations in their conjugations motivated by the adjunction of morphology units.
In the 1980s, Lexical Phonology gained prominence. The pioneers were Kiparsky
(1982, 1985) and Mohanan (1982), but many phonologists quickly adopted it because
it possesses high level of theoretical explanation, besides being responsible for making
morphology return to the setting of language studies (SCHWINDT, 2006).
LPM is dedicated to look at the lexicon not only as a keeper of structures, but also
as a set of phonological rules that communicate with morphological ones. One of its
greatest contributions is the understanding that the lexicon of the language is organized
in a series of levels or strata which are responsible for the occurrences of these rules.
Within each stratum, both morphological rules of word formation and phonological
processes are applied pairwise. Strata are arranged so as to reflect the ordering of the
word formation processes, as shown in Figure 1.

Alfa, São Paulo, v.62, n.1, p.169-189, 2018 171

 Figure 1 – Model proposed by Kiparsky (1982):
representation of the lexicon structure.

Source: Lee (1992).

The model shown in Figure (1), which was introduced by Kiparsky (1982),
considers the entry of the lexicon by the phonology module. From that point on, relations
between phonological and morphological strata begin to be established. Finally, the
word leaves the lexical level through phonology and goes directly to the syntax, at the
post-lexical level.
Lee (1995) emphasizes that the components of phonology and morphology are
mixed, so that relevant phonological rules apply to the output of any morphological
rule, determining entry to another phonological rule, and so on.
In the analysis of Portuguese, Bisol (2010) proposes the division of lexicon into two
levels, i. e., the root and the word, identified as Level 1 and Level 2, respectively. This
structure, shown in Figure (2), is the most used one in most LPM proposals, although
it is the option of each language to divide lexicon into different levels.

Figure 2 – Model suggested by Bisol (2010) for Portuguese.

Source: Bisol (2010, p. 92).

172 Alfa, São Paulo, v.62, n.1, p.169-189, 2018

 Bisol’s proposal (2010) introduces the LPM model with two possibilities from
the underlying structure. Input can happen either through the morphology module or
through the phonology module whereas the output of the word from the lexical module
to the post-lexical module is provided directly by phonology and follows the path to
the syntax. In this level, it undergoes post-lexical phonological rules.
In order to analyze the data of this study, its results are formalized with the support
of LPM, considering essentially Lee’s proposal (1995) for BP. On this basis, this study
advocates that the process of regularization of verb forms in the speech of 6-9-year-old
Brazilian children is due to the lack of correspondence between Phonology strata and
Morphology strata.

Methodology

The study was carried out with a group of 6-9-year-old native BP speakers – from
Rio Grande do Sul (RS) state, Brazil, – who speak a dialect variety that includes the
use of the personal pronoun ‘tu’ (‘you’). It focused on the production of irregular verbs
and aimed to identify whether the process of regularization of such verbs can still be
found in this age group. When it occurred, the study investigated the motivation of the
phenomenon within the assumptions of Lexical Morphology and Phonology.
The corpus consisted of data on eight native BP speaking children. All were literate
and monolingual and belonged to one of the age groups listed in Table 1. In addition,
children’s ages were linked to their schooling level. Table 1 shows the selection of
informants regarding schooling level and age1.

Table 1 – Subjects’ age groups and schooling levels.

Age Group 1 (FE 1): 06 years old 2 subjects 2nd year - Elementary school I
Age Group 2 (FE 2): 07 years old 2 subjects 3rd year - Elementary School I
Age Group 3 (FE 3): 08 years old 2 subjects 4th year - Elementary school I
Age Group 4 (FE 4): 09 years old 2 subjects 5th year - Elementary school I

Source: Goulart (2015).

In each group of selected ages, children were subdivided with respect to sex, i.
e., a girl and a boy composed each age group, in order to be able to establish relations
between the gender and possible peculiarities and/or similarities in the process of
acquisition of the irregular verb morphology of BP.

1
Because this paper is a part of a Master’s thesis, it has only some data of the research corpus and the description of only
one of the instruments created for data collection.

Alfa, São Paulo, v.62, n.1, p.169-189, 2018 173

 To collect data, an instrument, which focused on the production of irregular verb
morphemes in the 1st and 2nd singular persons of the Indicative Present, Simple Past
and Subjunctive Present, was created. The instrument consisted of sentences with gaps
to be completed by the informants with appropriate forms of the verb highlighted in
red, as shown in Figure 3.
All sentences led to the production of the expected verb tense. The test was presented
on slides, on the screen of a laptop (model shown in Figure 3). Two types of slides were
exhibited: one that highlights the verb to be conjugated and another one with sentences
whose gaps had to be completed with the inflected verb.

Figure 3 – Model of Instrument I2.

Source: Goulart (2015).

Initially, the researcher showed the first slide, containing the sentence with the verb
to be used. She asked the child to read the sentence and then, moving to the next slide,
asked the child to read and complete the sentences. Throughout the application, the
researcher tried to interact constantly with the informants. Preliminarily, a familiarization
test, similar to that of the instrument, had been applied, but with sentences containing
only regular verbs.

2
“To measure height is cool”.
I always ___ my height.
I ___ my height yesterday.
Daddy wants me to___ my height.
You___ your height everyday.
You___ your height yesterday.
Daddy wants you to___ your height.

174 Alfa, São Paulo, v.62, n.1, p.169-189, 2018

 The instrument also included slides that aimed at distracting the children, i. e.,
they had sentences with regular verbs. In addition, it should be emphasized that the
disposition of sentences in the test was randomly done to avoid the sequence of verb
conjugation in the order of grammatical persons, such as: eu faço, tu fazes; que eu
faça, que tu faças (‘I do’, ‘you do’; ‘me to do’, ‘you to do’). When, in the ordering of
sentences, the sequence of pronouns eu - tu (‘I - you’) appeared, the verb tense was
alternated.
The instrument had 10 slides which displayed irregular verbs. Each one had 06
different sentences with gaps to be filled. Thus, each informant had to fill 60 spaces
with the oral use of irregular verbs. Regular verbs, whose presence in the instrument
only occurred as a distraction from the focus of the research, totalized 48 regular verb
productions, distributed in 08 slides. Table 2 shows the irregular verbs selected for the
research proposed in this paper.

Table 2 – Irregular verbs and their alternations analyzed by the Master’s thesis.

Consonantal Phonological
Irregular Verbs
Alternation
Dizer (´to say’) [z], [g], [s]
Fazer (´to do’) [z], [s]
Satisfazer (´to satisfy’) [z], [s]
Trazer (´to bring’) [z], [g], [s]
Poder (´can’) [d], [s]
Ter 3 (´to have’) [], [v]
Perder (´to lose’) [d], [k]
Medir (´to measure’) [d], [s]
Ouvir (´to hear’) [v], [s]
Pedir (´to ask’) [d], [s]

Source: Goulart (2015). 3

In irregular verbs, whether by root change or by fugue to the paradigm, phonological

alternations can occur (shown in Table 2). They are characterized by replacing a
phoneme with another one. The principle of phonological alternations is maintained
by this study, since one can observe the replacement of a consonant with another one.
For example, in medir ‘(to measure’), one of the possibilities of conjugation is the
replacement of consonant “d” with consonant “s”. This fact occurs in the root of the
verb and, due to this fact, a phenomenon of morphological order is also manifested. In
view of this fact, LPM was brought to the analysis since it is a theory that can explain
linguistic processes that deal with units of phonology and morphology at the same time.

3
Consonant alternations related to the verb “ter” are of a differentiated nature in this list of verbs, since they refer to the
inflection paradigm, whereas alternations recorded for the other verbs occur in the root of the word.

Alfa, São Paulo, v.62, n.1, p.169-189, 2018 175

 According to Câmara Júnior (1970), consonantal alternation is constituted by
variation of the root, which contributes to express grammatical notions of tense, mood
and person that are primarily represented by suffixes. Souza, Silva and Koch (2009)
state that it is this type of irregularity that allows to distinguish deviant morphological
“patterns”, since one of the characteristics of regular verbs is root immutability.

Results and Data Analysis

In order to carry out data description, descriptive tables with information about
every informant were outlined. It should be emphasized that, at that moment, standard
and nonstandard forms, according to the BP irregular verb conjugation, were considered.
Therefore, the focus is to identify nonstandard forms, which are manifested through the
regularization process and do not show consonantal alternations which are necessary for
the inflections of irregular verbs, especially in the Indicative Present and Subjunctive
Present.
Thus, it should be pointed out that the consonantal alternation corresponds to the
form which, in this paper, is referred to as Standard (P)4, whereas its non-occurrence,
which corresponds to the production of the regularized verb form, is interpreted
as Nonstandard (NP). When NP productions are shown in the tables, the phonetic
transcription of the consonantal segment is also identified. It evidences the use of the
verb form in disagreement with the target language, that is, it implies the escape of the
morphophonological alternation displayed by the verb conjugation which is considered
standard5.

Informant 1
Informant 1/ FE 1/ F1 dizer fazer satisfazer trazer poder perder ter medir ouvir pedir

Indicative 1st person P NP NP NP P NP P NP NP NP

Present fa[z]o satisfa[z]o tra[z]o per[d]o mε[d]o ou[v]o pε[d]o
2nd person P P P P P P P P P P
Indicative 1st person P P P NP P P P P P P
Simple Past tra[z]i
2nd person P P P NP P P P P P P
tra[z]eu
Subjunctive 1st person P NP NP P P NP P NP NP NP
Present fa[z]a Satisfa[z]a per[d]a mi[d]a ou[v]a pε[d]a
2nd person P P NP P P NP P NP NP NP
Satisfa[z]a per[d]a mi[d]a ou[v]a Pe[d]a
vogal
média
baixa

4
The forms produced without morpheme-s, 2nd singular person marker (tu) were considered correct, due to the fact that
these manifestations are found in the frequent use of the language, even in the case of speakers with high educational
level. Thus, variants of the standard manifestation were the forms tu me[d]es ~ tu me[d]e, for example, since the focus of
the description fell on the consonant that is the target of morphophonological alternation in the verb forms under study.
5
Abbreviations kept the initials of the words in Portuguese: FE (age group); F and M (female and male); P (standard
form, that is, conjugated in agreement with the standard norm of the language); NP (non-standard form, i. e., conjugated
in a regular manner).

176 Alfa, São Paulo, v.62, n.1, p.169-189, 2018

 Informant 1, 06 years of age, female, presented regularizations in the conjugations
of the irregular verbs “fazer, satisfazer, trazer, perder, medir, ouvir and pedir”. Regarding
the verb “fazer” (‘to do’) and its derivative “satisfazer” (‘to satisfy’), the informant
produced regularized morphemes, which were faithful to the root of the verb faz-; as a
result, conjugations fa[z]o, fa[z]a, satisfa[z]o and satisfa[z]a were observed. Thus, the
girl did not perform the irregular conjugation, which would occur for the production of
alternations with [s], found in fa[s]o, fa[s]a, fi[s], both in the Indicative and Subjunctive.
Regularizations, which also maintain faithfulness to the base-morpheme of the verb,
such as tra[z]o, tra[z]i, tra[z]eu, were some of the conjugations that this informant
carried out. Thus, the /z/ phoneme alternated neither with [g] nor with [s] in the Indicative
mood, but alternated with [g] in the Subjunctive mood. Because the Subjunctive mood
is less frequent, the standard use may be interpreted as the result of an unanalyzed
form. Construction of the forms per[d]o and per[d]a, rather than per[k]o and per[k]
a, was also found.
The expected alternation for the verb “ouvir” (‘to hear’), of form [v] with [s], was
not performed by the child. On the contrary, what prevailed was the regularization
and faithfulness to the base form of the root ouv-. Such phenomenon was observed in
ou[v]o and ou[v]a, examples of the informant’s production.

Informant 2
Informant 2/ FE1/ M1 dizer fazer satisfazer trazer poder perder ter medir ouvir pedir

Indicative 1st person P P NP P P NP P NP NP NP

Present satisf[eito] per[d]o mi[d]o Ou[v]o pε[d]o
2nd person P P P P P P P P P P
Indicative 1st person P P P P P P P P P P
Simple Past
2nd person P P P P P P P P P P
Subjunctive 1st person P P P P P NP P NP NP P
Present per[d]a mi[d]a ou[v]a
2nd person P P P P P NP P NP NP P
per[d]a mi[d]a o[v]a

Informant 2, 06 years of age, male, conjugated the verbs “dizer, fazer, trazer,
poder and ter” in agreement with the standard grammar of the language, alternating
consonantal phonemes. However, in relation to the verbs “satisfazer, perder, medir,
ouvir and pedir”, the child produced regularized conjugations.
Concerning conjugations of the verb “perder” (‘to lose’), per[d]o and per[d]a were
produced. They are forms which were regularized by the child in both Indicative and
Subjunctive. As for the verb “ouvir” (‘to hear’), the forms ou[v]o, ou[v]a and o[v]a
were constructed. For the verb “medir” (‘to measure’), Informant 2 used mi[d]o and
mi[d]a, with vowel reduction [e → i]. With regard to the verb “pedir” (‘to ask’), the child
operated with the characteristic alternations of the Subjunctive Present and produced
pε[d]o in the Indicative Present, a regularization that underwent the process of vowel
alternation, but without effecting the consonantal alternation referring to /d/ for [s].

Alfa, São Paulo, v.62, n.1, p.169-189, 2018 177

 Informants 3, 4, 5, 6, 7 and 8, whose data are shown below, produced regularized
verb forms that show the same nature as those previously mentioned.

Informant 3
Informant 3/ FE2/ F1 dizer fazer satisfazer trazer poder perder ter medir ouvir pedir

Indicative 1st person P P NP NP P NP P NP NP NP

Present satisfa[z]o tra[z] per[d]o mε[d]o ou[v]o pε[d]o
2nd person P P P P P P P P P P
Indicative 1st person P P NP P P P P P P P
Simple Past satisfa[z]i
2nd person P P NP P P P P P P P
satisfa[z]
esse
Subjunctive 1st person P P NP NP P NP P NP NP NP
Present satisfa[z]a tra[z]a per[d]a mε[d]a ou[v]a pε[d]a
2nd person P P P NP P NP P NP NP NP
tra[z]a per[d]a mi[d]a ou[v]a pε[d]a

Informant 3, 07 years of age, female, made regularizations for the verbs “satisfazer,
trazer, poder, perder, medir, ouvir and pedir. The verbs “dizer, fazer and ter” were the
only ones that did not undergo this phenomenon.

Informant 4
Informant 4/ FE2/ M1 dizer fazer satisfazer trazer poder perder ter medir ouvir pedir

Indicative 1st person P P NP NP P NP P P NP P

Present satisfa[z]o tra[z]o per[d]o Ou[v]o
2nd person P P P P P P P P P P
Indicative 1st person P P P P P P P P P P
Simple Past 2nd person P P P P P P P P P P
Subjunctive 1st person P P NP NP NP NP P NP NP NP
Present satisfa[z]a tra[z]a po[d]a per[d]a mi[d]a ou[v]a pi[d]a
2nd person P P NP NP P NP P NP NP NP
satisfa[z]a tra[z]a per[d]a me[d]a ou[v]a pe[d]a

Informant 4, 07 years of age, male, had the regularization of verbal morphemes for
the verbs “satisfazer, trazer, poder, perder, medir, ouvir and pedir” in his productions.

Informant 5
Informant 5/ FE3/ F dizer fazer satisfazer trazer poder perder ter medir ouvir pedir

Indicative 1st person P P P P P P P P NP P

Present o[v]o
2nd person P P P P P P P P P P
Indicative 1st person P P P P P P P P P P
Simple Past 2nd person P P P P P P P P P P
Subjunctive 1st person P P P P P P P NP NP NP
Present mi[d]a o[v]a pi[d]a
2nd person P P P P P P P NP NP NP
me[d]a ou[v]a pe[d]a

178 Alfa, São Paulo, v.62, n.1, p.169-189, 2018

 Informant 5, 08 years of age, female, did not apply the morphophonological rules
that the irregular conjugations of the verbs “medir, ouvir and pedir” require.

Informant 6
Informant 6/ FE3/ M dizer fazer satisfazer trazer poder perder ter medir ouvir pedir

Indicative 1st person P NP NP P P NP P NP NP NP

Present fa[z]o satisfa[z]o per[d]o mε[d]o ou[v]o pε[d]o
2nd person P P NP P P P P P P P
satisfa[z]eu
Indicative 1st person P P NP P P P P P P P
Simple Past satisfa[z]i
2nd person P P NP NP P P P P P P
satisfa[z]esse tra[z]i
Subjunctive 1st person P NP NP P P P P NP NP NP
Present fa[z]a satisfa[z]a mi[d]a ou[v]a pε[d]a
2nd person P NP NP P P P P NP NP NP
fa[z]a satisfa[z]a me[d]a ou[v]e pi[d]a

Informant 6, 08 years of age, male, only conjugated the verbs “dizer, poder and
ter” in agreement with the language pattern.
As for the other verbs of the study, Informant 6 distanced himself from the pattern
in the production of some inflections and produced them in regularized form, without
the use of the morphophonological rule. Data on Informant 6, as well as data on the
other children, indicate that the child, who is still in the phase of acquisition of the
morphophonology of the language, reveals floating morphological characteristics by
comparison with the standard grammar of the target language. These are features that
may either show up properly or may be far from the pattern.

Informant 7
Informant 7/ FE4/ F dizer fazer satisfazer trazer poder perder ter medir ouvir pedir

Instrument Indicative 1st P P P P P P P P P P

I Present person
2nd P P P P P P P P P P
person
Indicative 1st P P P P P P P P P P
Simple person
Past 2nd P P P P P P P P P P
person
Subjunctive 1st P P P P P P P NP P P
Present person mi[d]a
2nd P P P P P P P NP P P
person mε[d]a

Informant 7, 09 years of age, female, used regularized verb conjugations for the
verb “medir” (‘to measure’) in the Present Subjunctive; both forms mi[d]a and mε[d]a
show that, although the child’s grammar is almost reaching the target system of the
standard language, the instability generated by the complexity of the acquisition of BP

Alfa, São Paulo, v.62, n.1, p.169-189, 2018 179

morphology is still present at the age of 09. They are conjugations that are produced
in regularized form, especially in the inflections in the Subjunctive mood of the verb
system.

Informant 8
Informant 8/ FE4/ M dizer fazer satisfazer trazer poder perder ter medir ouvir pedir

Instrument Indicative 1st P P NP P P P P NP NP NP

I Present person satisfa[z]o mε[d]o o[v]o pε[d]a

2nd P P P P P P P P P P
person
Indicative 1st P P P P P P P P P P
Simple person
Past
2nd P P P P P P P P P P
person
Subjunctive 1st P P NP P P P P NP NP P
Present person satisfa[z]a mi[d]a o[v]a

2nd P P NP P P P P NP NP P
person satisfa[z]a mε[d]a o[v]a

Informant 8, 09 years of age, male, produced inflections whose roots did not comply
with the rules of consonantal alternations required by the irregular Portuguese verb
pattern for the verbs satisfazer, medir, ouvir and pedir.
Like the other children in this study, this informant regularized some conjugations –
that pertain to the standard of PB regular verbs –, guided by the root of the irregular
verb in the infinitive. The forms satisfa[z]o, satisfa[z]a, m[εd]o, m[id]a, m[εd]a, o[v]o,
o[v]a and p[εd]a are examples of this regularization.

Data in the light of Lexical Phonology and Morphology

All informants performed some verb inflection in disagreement with the standard
norm of conjugation of the irregular verb category. Table 3 shows how often
regularizations of each verb under investigation were produced by the children.

180 Alfa, São Paulo, v.62, n.1, p.169-189, 2018

 Table 3 – Estimates of the regularization process.

FORMS OF NONSTANDARD (NP) PRODUCTION

The use of verb forms
Instrument I
Verb
Number of
Production
occurrences
DIZER 100% production in agreement with P (standard)
02 fa[z]o
FAZER
03 fa[z]a
05 satisfa[z]o
01 satisfa[z]ei
01 satisfa[z]eu
SATISFAZER 01 satisf[eito]
02 satisfa[z]i
03 satisfa[z]esse
11 satisfa[z]a
05 tra[z]o
04 tra[z]i
02 tra[z]eu
TRAZER
01 tra[z]
01 tra[z]ei
07 tra[z]a
PODER 01 po[d]a
10 per[d]o
PERDER
16 per[d]a
TER 100% production in agreement with P (standard)
06 mε[d]o
01 me[d]o
03 mi[d]o
MEDIR
19 mi[d]a
05 mε[d]a
06 me[d]a
07 ou[v]o
02 ou[z]o
04 o[v]o
OUVIR
18 ou[v]a
03 ou[z]a
05 o[v]a
06 pε[d]o
01 pi[d]o
PEDIR 05 pe[d]a
07 pε[d]a
05 pi[d]a
Source: Goulart (2015).

Alfa, São Paulo, v.62, n.1, p.169-189, 2018 181

 Data show that the most recurrent regularizations in the informants’ conjugations
were forms per[d]o, per[d]a (verb - perder), m[id]a (verb - medir), p[εd]a, p[id]a
(verb - pedir), ou[v]a (verb - ouvir) and satisfa[z]a (verb –satisfazer). The Subjunctive
mood (Subjunctive Present) has greater instability as to its inflection, in high percentage
by comparison with productions in the Indicative Present. This fact is believed to be
related to the complexity that involves the acquisition of the morphophonology of BP and
to the fact that it is the tense and the verb mood (Subjunctive Present) that is considered
marked in the language. The more marked the verb mood in the language, the greater
the possibility of variation in relation to the standard form. The Subjunctive mood is
also used in subordinate sentences; it implies the construction of a period composed
of subordination and requires great syntactic complexity to the point that, in the oral
use of the language, it is seldom used, even by adults.
The Indicative Simple Past had low index of regularization, which is attributed
to the fact that this is a tense that demands few consonant changes in its inflections in
the Indicative mood. Moreover, its use does not require syntactic complexity either,
since it can either be employed in simple periods or composed of coordination, which
is often used in the language.
Regarding the Indicative Present, data can be considered relevant, since they are
based on previous studies (YAVAS; CAMPOS, 1988; SANTOS; SCARPA, 2003) which
consider this tense as one of the first to be acquired by native BP speakers. Thus, it may
be inferred that, although it is a tense which is early used by the subjects, especially
in the regular class of verbs, the irregular verb forms of the Indicative Present that are
carried out with consonantal alternations are complex for children in the acquisition
phase, a fact that contributes to their late emergence6.
Difference in non-standard production between boys and girls was below 1%, a
fact that leads to the understanding that verb regularization is not related to the subjects’
gender. In view of it, it is evident that the use of morphophonological events in the
conjugation of irregular verbs is found in the linguistic development of all children
in the sample, regardless of the gender, unlike other language acquisition processes
in which girls and boys may present different behavior, such as the acquisition of the
segmental inventory of language phonology.
The corpus under analysis also reveals that there is gradual development of irregular
verb acquisition in relation to age, which is also superimposed on the subjects’ schooling
level. The older the group (and the higher the schooling level), the more the process of
producing different forms from the standard (NP) decreases.
This relationship proposes that irregular verb forms of BP are acquired as the child’s
exposure to the language system increases. So, the older the subjects, the lower the
regularization rate they produce.

6
Data described in this paper also point out that the phenomenon of verb regularization occurs at higher percentage
with verbs of the third conjugation (-ir). The determining factor that explains these occurrences was not investigated
in detail, but it is emphasized that instruments have been developed to investigate this influence, as well as to highlight
the possible differences that exist in irregular verb acquisition, when dealing with different verb moods.

182 Alfa, São Paulo, v.62, n.1, p.169-189, 2018

 The frequency of use of verbs was a factor that was of particular relevance to
this analysis. In this context, data show that non-consonantal alternation occurs, in a
prevalent way, when dealing with the less frequent verbs in the informant’ linguistic
input, as Andersen (2008) had already pointed out. He stated that the verbs “ter, poder,
dizer and fazer” are among the ten most frequent BP verbs. In fact, in this research,
these were the verbs that least suffered the process of verb regularization. Besides, both
verbs “ter and dizer” were not inflected any time in a nonstandard way, whereas the
verbs “poder and fazer” were targets of a very low number of regularized productions.
In accordance with Lexical Phonology and Morphology, there are at least three
explanations for the process that affects NP conjugations, that is, regularized conjugations
of verb forms, evidenced by this paper:

a) the assumption that the subjects have not yet fully acquired the Verbal
Morphology of the language and, because of it, present a blockade at the
morphological level;
b) the assumption that the lack of alternations in the conjugation of irregular roots
is motivated by a phonological gap, directly linked to the phonological level;
c) the assumption that the child has already acquired both the Verbal Morphology
of BP and its Phonology and, thus, has regularized verbal morphemes due to
the motivation required by the morphological relationship that, in the lexical
component, requires interaction between Morphology and Phonology.

Firstly, regarding (b): it does not seem to account for verb forms produced differently
from the standard, since the segmental and syllabic inventory of the child with typical
development and most rules that determine its functioning in the language are acquired
until the age of 05 (LAZZAROTTO-VOLCÃO, 2009).
Considering (a), it seems to be discarded, because data collected by this study show
that the informants have already acquired the Verbal Morphology, since they master, in
accordance with the language pattern, the morphemes of tense and mood and of number
and person, that is, all verb inflections, including the verbal paradigm. Examples of this
fact are the inflections satisfa[z]o and per[d]a, which reveal that informants retain, in
the different forms of the pattern, the tense-indicating morphemes, as well as those of
number and person. As a result, it could be understood that Verbal Morphology in all
age ranges under study has already been acquired. It seems, therefore, that inflectional
morphemes of the class of verbs, corresponding to the categories under investigation
by this study, have already been acquired. In addition, all children, in all age groups
under study, used the forms of the verb ter (‘to have’) at 100% adequacy.
The alternative proposed in (c) is able to explain the phenomenon of nonstandard
occurrences, demonstrating, through LPM (Figure 3), that regularizations are motivated
by the lack of correlation that must be maintained between Morphology and Phonology so
that the process of alternations in irregular verbs follows the pattern of the language. The
phonological rule of alternation is not applied to the conjugation of some verbs, because

Alfa, São Paulo, v.62, n.1, p.169-189, 2018 183

there is no adequate match between morphological and phonological components and,
therefore, the morphophonological interaction is not carried out.
Figure 4 shows lack of match between phonological and morphological components,
based on Kiparsky’s proposal (1982, 1985), in order to understand the forms produced
by the subjects in this research.

Figure 4 – Proposal for nonstandard forms, based on Kiparsky (1982, 1985).

Source: Goulart (2015), based on Kiparsky’s Model (1982).

In the theoretical model of LPM, proposed by Lee (1995)7 for Brazilian Portuguese,
irregular flexion is at level 1 (α), while regular flexion manifests at level 2 (β).
Verb productions carried out by children in this study indicate that, at the age of
06, subjects already operate with irregular forms, since they adequately produce the
consonant phonological alternations of certain verbs, especially those whose frequency
is high in the linguistic input.
However, data also point out that, in all age groups under study, children still regularize
irregular verb forms, thus evidencing the late acquisition of morphophonological
processes of the BP verb system, especially of less frequent verbs (the case previously
emphasized) and the less frequent verb tense: Subjunctive Present. Simultaneous use
of irregular forms for some verbs and non-use for others lead to the interpretation that
the acquisition of these verbs requires incorporation into the child’s grammar:

I) of regular inflectional verbal morphemes (mood and tense, number and person);
II) of irregular verbal morphemes (irregularities due to the paradigm);
III) of alternations of irregular roots (irregularities due to root).

Facts in (II) and (III) are irregular; therefore, they are unpredictable and have to
be learned by the child with the use (the frequency) of the language. Besides, they

7
According to Lee (1995), at level 1 (also identified as (α)), there are irregular inflections, the Derivation and the
Composition I; at level 2 (β), productive formation of PB occurs, together with regular verb inflections of the language;
in the post-lexical level, there is Composition II. For further details, see Lee (1995).

184 Alfa, São Paulo, v.62, n.1, p.169-189, 2018

require interaction between Morphology and Phonology. For this reason, the frequency
of the use of verbs in the language is a conditional factor of the acquisition of irregular
verbs, i. e., the more frequent the use of an irregular verb, the more easily its structure
will be acquired.
Based on these arguments, the proposal in (c) is the one that seems to be most
adequate to explain the corpus of this study, taking into account, above all, the
morphophonological nature of the process, which is essential for the standard irregular
BP. Because it is complex in language, it contributes to the late acquisition of verb
forms whose root involves consonantal alternations in their conjugations.
According to Lee’s proposal (2005), in the analysis of data described by this paper,
non-communication of morphological and phonological components of the language
occurs only when considering irregular verb forms, which are inflected in the child’s
speech in a regularized way. Then, they go to level 2 (β) without undergoing rules of
irregular inflection and, at level 2 (β), they are submitted to the treatment given to
regular verbs. This division of verb, irregular and regular inflection in BP at different
levels occurs in compliance with the Elsewhere Condition Principle, in which more
restricted rules apply before the most general ones, that is, irregular inflection, which
is the most marked and most restricted one, occurs before regular inflection, whose
rules are less marked and more universal, not only in BP, but in languages of the world.
In the passage to level 2 (β), the verb that did not undergo irregular inflection
(because, despite submitting to morphological rules, it did not reach the consonantal
changes at the phonological level8 due to lack of correlation between the components)
suffers regular inflection exclusive of level 2 (β), preserving itself in a regular and
faithful way to the root of the verb, which is a morpheme that is in the deep lexicon,
integrating the underlying representations of the speakers of the language.
According to Lee (1995), it is also at level 2 (β) that the productive formation
of BP occurs, together with regular verb inflections of the language. It highlights the
reason why regular verbs are understood as the productive class of Portuguese and,
therefore, of unmarked nature in the language. In this sense, regularization in the
process of verb acquisition shows that children operate in accordance with the unmarked
pattern. It becomes the most natural phenomenon, when they have not yet acquired the
irregularities of the linguistic system.
In addition, this fact may also show that, when speakers produce regularizations
(nonstandard forms), they are dealing with a single internalized grammar, unlike studies
(LORANDI, 2006; AVELEDO, 2006) which postulate two operating grammars in the
verb system: one is responsible for the conjugation of regular verbs and the other one
is in charge of the conjugation of irregular verbs. Regularization, in these researches,
is understood as being an inflection established by the grammar of regular verbs and,
consequently, maintained in the subjects’ speech with faithfulness to the root of the verb.

8
The locus of the absence of consonantal alternation, as part of the process of acquisition of irregular verb forms, is
represented in Figure 4 (on the right).

Alfa, São Paulo, v.62, n.1, p.169-189, 2018 185

 In the light of the Lexical Theory, the process of non-standard forms was observed
to occur through the relations established between the components and strata that
constitute the grammar of the language. Thus, the regularization phenomenon is situated
in the field that establishes the interchange between the components to form the BP
vocabulary and, because of it, it is a process that operates with a single grammar, capable
of explaining both the standard and the non-standard inflection. It seeks for linguistic
parameters for events that mark the development of language acquisition, specifically
the acquisition of BP irregular verbs.

Final considerations

In the verb acquisition phase, the child operates with at least two possibilities
of inflection of irregular verbs: either in agreement with the target of the language,
performing phonological alternations in line with the pattern; or in disagreement with the
target, carrying out the phenomenon that is understood in the literature as regularization
(LORANDI, 2006). Inflection that occurs without the proper alternations of irregular
conjugation maintains the root of the verb in the subject’s speech and, therefore, inflects
the irregular class in accordance with the conjugation of regular verbs.
Taking into account that Brazilian studies carried out up to the present moment attest
that the phenomenon of verb regularization just occurs until the age of 5 years, data
collected by this research disagree, since it shows evidence that children over the age
of 5 years produce regularized verb forms in the conjugation of certain irregular verbs.
Therefore, this study reaffirms the importance of the schooling level, a variable
linked to age, to the stabilization of BP irregular inflection. In addition, the analysis
concluded that:

a) the corpus of this study shows, through the conjugations under analysis, that the
subjects under investigation have already got the morphology of the language,
since they produce the morphemes of mood and tense, of number and person
and the markers of the BP verb paradigm;
b) regularizations in the conjugation of irregular verbs in the view of this study,
which is based on LPM, have morphological nature in the BP system, because
they are associated with relations conditioned by the mismatch between the
phonological and the morphological components of the language;
c) because it implies a morphophonological phenomenon, acquisition of the
irregular verb system is late and may extend to at least 9 years of age, in
the case of native BP speakers; in this study, the variable age, linked to the
schooling level, points out that the greater the child’s linguistic experience,
the greater the irregular conjugations s/he produces, considering the standard
form of the language;
d) verbs – and their inflections – that are maintained at high frequency in the
language are acquired earlier than those that show low frequency.

186 Alfa, São Paulo, v.62, n.1, p.169-189, 2018

 Under the assumptions of Lexical Phonology and Morphology, the use of verb forms
which are different from the standard, without the presence of consonantal alternations,
can be explained by the non-articulation between the modules of Morphology and
Phonology of the language. Lack of matching between these components in the
construction of certain irregular conjugations, especially in the first person of the
Indicative Present and in the 1st and 2nd persons of the Subjunctive Present, points
to the high complexity that permeates the acquisition of the BP irregular verb class.
Considering that the subjects of this study already operate with this relation and produce
consonantal alternations, mainly for the high frequency verbs in the linguistic input, this
study understands that they are still in phase of acquisition of the morphophonology
of the language due to the coexistence of standard (P) and nonstandard forms (NP) in
their productions.
With this analysis, we have evidence of the complexity of morphophonological
phenomena and, in this fact, there is a justification for the late stage of its acquisition: the
acquisition of morphology of verbs is complex due to the different affixes that it involves.
Besides, difficulty gradually increases when tenses are in the subjunctive mood (from
the syntactic and semantic point of view, the use of this verb mood implies a demanding
operation in the relations between clauses and meanings). Complexity becomes even
greater when there is interaction between the morphological and phonological levels
of the language, thus, requiring vowel and consonantal alternations in the process of
verb inflection.
Results achieved by this study therefore confirm that, when phenomena of
morphological nature are considered, the process of language acquisition by children
can extend to the age of 9 years or more, including the use of irregular verbs.

GOULART, T.; MATZENAUER, C. A conjugação de verbos irregulares por crianças falantes

nativas de português brasileiro: um estudo sob o viés da fonologia e morfologia lexical. Alfa,
São Paulo, v.62, n.1, p.169-189, 2018.

■■ RESUMO: Este estudo investiga o processo de aquisição de verbos irregulares do Português

Brasileiro (PB), quando conjugados nos tempos Presente do Indicativo, Presente do Subjuntivo
e Pretérito Perfeito do Indicativo, por crianças falantes nativas da língua, com o objetivo de
descrever e analisar, sob o enfoque da Fonologia e Morfologia Lexical (Lexical Phonology and
Morphology - LPM), a relação morfofonológica presente nas flexões dos verbos irregulares.
Os dados revelam que os fenômenos linguísticos de origem morfofonológica são de alta
complexidade para os falantes e podem ser considerados como processos de aquisição tardia
pelas crianças brasileiras. Por meio da LPM compreendemos que a regularização verbal, uma
das manifestações das conjugações irregulares, pode ser explicada pela não-correspondência
entre os níveis fonológico e morfológico da língua.

■■ PALAVRAS-CHAVE: Aquisição morfofonológica. Verbos irregulars. Fonologia e Morfologia

Lexical.

Alfa, São Paulo, v.62, n.1, p.169-189, 2018 187

REFERENCES

ANDERSEN, E. M. L. Representações Lexicais Subjacentes: verbos e léxico inicial.

ReVEL, v.6, n.1, p.1-31, 2008.
AVELEDO, F. El processamento de verbos regulares e irregulares en el español infantil:
mecanismo dual vs. Conexionismo. Boletín de Linguística, v.18, n.26, p.5-32, jul. 2006.
BASILIO, M. Teoria Lexical. 7. ed. São Paulo: Afiliada, 2004.
BISOL, L. Introdução e estudos de fonologia do português brasileiro. 5. ed. Porto
Alegre: EDIPUCRS, 2010.
CÂMARA JÚNIOR, J. M. Uma forma verbal portuguesa: estudo estilístico-
gramatical. Rio de Janeiro: Jornal do Comércio – Rodrigues & Cia, 1975.
_______. Estrutura da Língua Portuguesa. 38. ed. Rio de Janeiro: Vozes, 1970.
CASTILHO, A. T. de. Nova Gramática do Português Brasileiro. São Paulo: Contexto,
2012.
GOULART, T. P. D. A Produção de Formas Verbais Irregulares por Crianças
Falantes do Português Brasileiro (PB). 2015. 147 f. Dissertação (Mestrado em
Linguística Aplicada) – Universidade Católica de Pelotas, Pelotas, 2015.
KIPARSKY, P. Lexical morphology and phonology. In: YANG, S. (Ed.). Linguistic
in the morning Calm. Seoul: Hanshin Publishing Co, 1982.
________. Some consequences of lexical phonology. Phonology yearbook 2. London:
Cambridge University Press, 1985.
LAZZAROTTO-VOLCÃO, C. Modelo Padrão de Aquisição de Contrastes: uma
proposta de avaliação e classificação dos Desvios Fonológicos. 2009. 218 f. Tese
(Doutorado em Letras) – Universidade Católica de Pelotas, Pelotas, 2009.
LEE, S. Fonologia Lexical do Português. Cadernos de Estudos Linguísticos,
Campinas, v.23, p.103-120, 1992.
_______. Morfologia e Fonologia Lexical do Português do Brasil. 1995. 201 f.
Tese (Doutorado em Linguística) – Instituto de Estudos da Linguagem, Universidade
Estadual de Campinas, Campinas, 1995.
_______. Fonologia Lexical – Modelos e Princípios. Letras de Hoje, Porto Alegre,
v.31, n.2, p.129-137, 1996.
LORANDI, A. Formas Morfológicas na Gramática Infantil: um estudo à luz da
Teoria da Otimidade. 2006. 185 f. Dissertação (Mestrado em Linguística Aplicada) –
Pontifícia Universidade Católica, Porto Alegre, 2006.

188 Alfa, São Paulo, v.62, n.1, p.169-189, 2018

MOHANAN, K. P. Lexical Phonology. 1982. 218 f. Doctoral dissertation –
Massachusetts Institute of Technology, Cambridge, 1982.
SANTOS, R. S.; SCARPA, E. M. A aquisição da morfologia verbal e sua relação com
o acento primário. Letras de Hoje, Porto Alegre: EDIPUC, v.38, n.4, p.249-260, 2003.
SCHWINDT, L. C. A relação entre Morfologia e Fonologia na História dos Estudos
Linguísticos. In: MARTINS, E. S.; CANO, W. M.; MORAES FILHO, W. B. (Org.).
Léxico e morfofonologia: perspectivas e análises. Uberlândia: EDUFU, 2006.
SOUZA & SILVA, M. C. P.; KOCH, I. V. Linguística Aplicada ao Português:
Morfologia. 17. ed. São Paulo: Cortez, 2009.
WETZELS, W. L. Mid Vowel Neutralization in Brazilian Portuguese. Cadernos de
Estudos Linguísticos, n.23, p.19-55, 1992.
YAVAS, F.; CAMPOS, J. Aquisição da Morfologia Verbal do Português como L1 e L2.
Letras de Hoje, Porto Alegre: PUCRS, n.74, p.81-95, 1988.

Received in March 27, 2017

Approved in September 1, 2017

Alfa, São Paulo, v.62, n.1, p.169-189, 2018 189