Φ Revista de Filología y Lingüística de la Universidad de Costa RicaPublicación Semestral, ISSN-0377-628XVolumen 40 - Número 1Enero - Junio 2014
ChiLdRen’s Use oF spanish taps: a 
natURaListiC stUdy with monoLingUaL
Costa RiCan ChiLdRen ages 3;0 to 5;6
Luz Marina Vásquez Carranza
Esta obra está bajo una licencia Creative Commons 
Reconocimiento-No Comercial-Sin Obra Derivada

ResUmen
Este estudio reporta las tendencias en el uso de la vibrante simple en 15 niños(as) monolingües de 
habla hispana costarricenses con edades entre los 3;0 y los 5;6. Los datos obtenidos por medio de 
grabaciones de audio del lenguaje natural de estos niños(as) produjeron 1080 palabras claves. El 
80% de ellas fueron usos correctos de la vibrante simple, mientras que el 20% restante se trató de 
omisiones (14%) o sustituciones [l], [n], [t], [d], [j], [tS], [D] o por una rótica asibilada no vibrante. 
Un hallazgo principal de esta investigación se refiere a la omisión de la vibrante simple en verbos 
infinitivos seguidos de uno o más pronombres enclíticos. Este tipo de omisión es interesante en 
cuanto parece estar directamente relacionado con la complejidad de los pronombres enclíticos, ya 
que los(as) niños(as) no siempre omiten la vibrante simple en contextos similares (/Cr/). Aunque 
este patrón ha sido reportado con respecto al español de los adultos costarricenses, este nunca ha 
sido reportado en el habla infantil.
palabras clave: róticas del español, la vibrante simple del español, desarrollo del lenguaje infantil, 
patrones en la adquisición del lenguaje.
aBstRaCt
This study reports on the trends in children’s use of taps by 15 Costa Rican Spanish-speaking 
monolingual children between ages 3;0 and 5;6. Data were obtained through audio-recordings of 
the children’s naturalistic speech, yielding 1080 target-words. 80% of the taps were correct, but 
the remaining 20% evidenced either omissions of the tap (14%) or substitution of the tap for [l], 
[n], [t], [d], [j], [tS], [D], or for a voiceless assibilated rhotic. A main finding in this study was the 
consistent omission of the tap in infinitive verb forms before a consonant initial enclitic pronoun. 
This omission type is interesting in that it appears to be directly related to the complexity involved 
in nominal enclitics, as children do not always omit the tap in similar phonetic contexts (i.e., /Cr/ 
clusters). Although this last pattern has been reported for adult Costa Rican Spanish, it has never 
been reported in child speech.
Key words: Spanish rhotics, the Spanish tap, child language development, patterns in acquisition.
ChiLdRen’s Use oF spanish taps: a 
natURaListiC stUdy with monoLingUaL
Costa RiCan ChiLdRen ages 3;0 to 5;6
Luz Marina Vásquez Carranza
Filología y Lingüística 40 (1): 217-228, 2014 
issn: 0377-628X
ph.d. Luz marina Vásquez Carranza. Universidad de Costa Rica. Sede de Occidente. Directora del 
Departamento de Filosofía, Artes y Letras. Docente de Sección de Lenguas Modernas.
Correo electrónico: luzmarinave@hotmail.com
Recepción: 24- 03- 2014
Aceptación: 02- 06- 2014
eL Uso de La ViBRante simpLe o eRe en eL LengUaje inFantiL: 
Un estUdio Con niños y niñas monoLingUes deL españoL 
CostaRRiCense Con edades entRe Los 3 y Los 5;6
Filología y Lingüística 40 (1): 217-228, 2014/ ISSN: 0377-628X218
1. introduction
Spanish rhotics have been studied in detail, and it has been shown that they involve 
complex gesture that generally results in late acquisition (e.g., Gómez-Fernández, 2004; Jiménez, 
1987; Vásquez-Carranza, 2006); one such rhotic sound is the Spanish tap, []. Most likely as 
a result of articulatory complexity, taps reportedly evidence assimilation processes across 
different adult Spanish dialects (e.g., Bradley & Schmeiser, 2003; Umaña-Aguilar, 1981), and 
some of these patterns are also found in child speech (e.g., Gómez-Fernández, 2004).
This study was designed to observe the trends in children’s use of taps by a group of 
15 Costa Rican Spanish-speaking monolingual children between ages 3;0 and 5;6. The data 
were obtained through audio-recordings of the children’s naturalistic interactions with peers 
and with the main researcher.
One thousand and eighty (1080) examples of words requiring taps in all possible 
phonetic contexts were analyzed, and the majority of the examples accurately contained taps. 
Twenty per cent (20%) of the target words, however, evidenced either omissions of taps (14%) 
or some sort of substitution, including substitutions for [l], [n], [t], [d], [j], [t], [D], as well 
as for a voiceless assibilated rhotic. The first substitution has commonly been reported for 
child Spanish, whereas the other substitutions were found in only a few words. An interesting 
example of omission, one which has not previously been reported for child speech, consisted 
of omissions of the tap in infinitive verb forms before a consonant initial enclitic pronoun. 
This omission type is interesting in that it appears to be directly related to the complexity 
involved in nominal enclitics.
2. the tap among spanish rhotics
Spanish is one of the few languages in the world that has two rhotics, a trill and a 
tap; they overlap only intervocalically, as elsewhere they are either neutralized or in free 
variation (Bradley, 2005; Proctor, 2009). Specifically, the trill emerges in onset position 
word-initially (e.g., rey ‘king’), in onset position word-medially after the consonants [l], [n], 
and [s] (e.g., alrededor ‘around’; enredo ‘mess’; Israel), and intervocalically, spelled as -rr- 
(e.g., arroz ‘rice’). The tap, in contrast, is found in word-medial pre-consonantal codas –i.e., 
part of the syllable which comes after the nucleus− (e.g., carta ‘letter’), in word-final codas 
(e.g., comer ‘to eat’ / color), in onset /Cr/clusters (e.g., tres ‘three’, drástico ‘drastic’), and 
inter-vocalically (e.g., aros ‘rings’).
To articulate a trill requires very precise complex gesture, as it consists of a series 
of brief occlusions of the tongue tip against the alveolar ridge while maintaining constant 
vibration of the vocal folds (Ladefoged, 2001). In contrast, the tap is articulated by rising the 
tongue up to touch the alveolar ridge and then lowering it back to the bottom of the mouth; it is 
a sound that has extra short duration and which is phonetically described as a voiced alveolar 
tap (Ladefoged, 2001; Ladefoged & Madieson, 1996; Recasens, 1991).
3. Cross-dialectal phonetic realizations of the spanish tap
The Spanish tap undergoes diverse phonetic realizations across various dialects. First, 
in the Spanish spoken in some regions of Cuba (Santiago), Panama, Dominican Republic, 
and Puerto Rico, the tap is lateralized in coda position, in words such as verdad [vel.da] and 
comprar [kom.p|al] (López-Morales, 1992; Willis, 2006).
VÁSQUEZ: Children’s use of Spanish taps: A naturalistic study... 219
In some Spanish dialects in Colombia and the Canary Islands as well as in the 
Cibao region in the Dominican Republic, coda liquids tend to vocalize in pre-consonantal 
position as well as word-finally in words with final stress. In such contexts, the rhotic is 
phonetically realized as a high front vowel or as a palatal glide, as in cuerpo [kwej.po] and 
mujer [mu.hej] (Proctor, 2009).
In Standard Spanish, there is a vowel segment that intervenes in complex /Cr/ onsets 
(i.e., portion of a syllable preceding the nucleus) and which may have variable duration, as in 
pronto [pә|] ‘soon’ and fresco [fә|] ‘fresh’; this intervening segment is often referred to as a 
svarabhakti vowel (Bradley & Schmeiser, 2003). In contrast, in casual speech in Peninsular 
Spanish as well as in some contemporary American Spanish dialects, these two authors 
report coarticulation of complex onset clusters; that is, /Cr/ clusters are heavily overlapped 
and such coarticulation results in friction of the rhotic (the intervening vowel disappears). The 
rhotic hence tends to devoice after voiceless consonants such as [k], and dental [d 1] and [t 1] 
assimilate regressively to the rhotic, resulting in [t® 9], an alveolar quasi-affricate sound. Bradley 
& Schmeiser (2002) point out that whereas in casual speech in Peninsular and Peruvian 
Spanish coarticulation affects any /Cr/ cluster, in other Latin American dialects it is restricted 
to consonant clusters in which the first consonant is a coronal non-continuant (i.e., [t] or [d]). 
Bradley and Schmeiser add that the tap becomes an approximant in tr- onset clusters, as in tres 
(‘three’), resulting in a rhotic similar to that found in English words such as dress and train, 
and which is not typical of Spanish. Quilis (1999) also identifies an affricated realization of 
the tap in [t] onsets in great part of Argentina, in Bolivia, some areas of the United States, 
in Guatemala, and in some parts of Spain, namely along the Ebro River. He adds that this 
affrication phenomenon also applies to the [d|] sequence, which is articulated as a voiced post-
alveolar affricated segment. Stockwell & Bowen (1965) report this coarticulation phenomenon 
for parts of Chile and Perú.
On the other hand, Bradley (2005) reports that in Standard Spanish /rC/ clusters where 
the rhotic appears in coda position there is also evidence of a svarabhakti vowel segment 
between the rhotic and the following consonant segment, but in casual speech in Highland 
Ecuador, this intervening vowel segment often tends to disappear, and an assibilated [r #] 
surfaces instead before consonants that have the same voicing value (e.g., verde [er #.De] ‘green’; 
carne [kar #.ne] ‘meat’). Such an assibilated segment devoices before voiceless consonants, 
whereas in any other context, [] surfaces instead.
Coarticulation in /t|, d|/ clusters is also reported for Costa Rican Spanish. For instance, 
Umaña-Aguilar (1981) argues that after the voiceless alveolar stop [t], the tap receives an 
alveolar articulation very similar to a post alveolar voiceless affricate; this rhotic is said to be 
retroflex and to closely resemble the American English /t/. The same phenomenon is reported 
in Vásquez-Carranza (2006) after the coronal voiceless stop phoneme [t], where // devoices 
and assibilates both word-initially and word-medially, as in atrás /atr # 9as/ ‘behind’ and trato /
tr # 9ato/ ‘deal’. Vásquez-Carranza (2006) reports the same assibilation process for word-medial 
complex dr onset clusters after a voiced sonorant coronal (l or n), as in vendrá (‘he/she will 
come’) and saldré (‘I will go out’).
In syllable-final position, the tap is also said to assibilate in Costa Rican Spanish. 
Umaña-Aguilar (1981) describes a tap which is phonetically realized as an assibilated 
retroflex in syllable-final position, especially phrase-finally. Sánchez-Corrales (1986) and 
Vásquez-Carranza (2006) also argue for this pattern in Costa Rican Spanish. A word-final 
Filología y Lingüística 40 (1): 217-228, 2014/ ISSN: 0377-628X220
assibilated rhotic is also reported for Highland Ecuador in Bradley (2001). He describes 
an assibilated [r #] word-finally in pre-vocalic position, as in ir ahora ‘to go now’, before a 
consonant, as in mayor gusto ‘greatest pleasure’, and before a phrase boundary, as in la flor 
‘the flower’. Similarly, in a dialect spoken in Mexico City, Harris (1969) reports a voiceless 
fricative, which is different from the alveolar in that it is retracted, in word-final contexts in 
words such as tomar ‘to take/to drink’.
Calvo-Shadid & Portilla-Chaves (1998) describe three assibilated retroflex variations 
of the rhotics found in the formal speech of four female informants in the capital area of Costa 
Rica. In particular, they identified a voiced, retroflex, approximant [’], a voiced retroflex 
fricative [Ω], and a voiced retroflex tap. Inter-vocalically, 4% of the allophones of the tap 
were described as retroflex, and the retroflex allophones were found mainly in pre-consonant 
position. Their study did not report evidence of a retroflex segment in consonant clusters, as 
do Umaña-Aguilar (1981) and Vásquez-Carranza (2006).
Finally, according to Penny (1969; cited in Bradley, 2005), in Northern Peninsular 
Spanish (i.e., Cantabrian Spanish), the tap that appears in infinitive verbs is lost before a 
consonant-initial clitic (e.g., la/lo/las/se/me), as well as before definite articles heading a 
following noun phrase (e.g., comer los huevos ‘eat the eggs’). Bradley (1998) reports the same 
phenomenon for Ecuadorian Spanish. In the Spanish spoken throughout the Central Valley in 
Costa Rica, deletion of the final rhotic in infinite verbs before a nominal clitic has also been 
reported. Specifically, Vásquez-Carranza (2006) states that the tap tends to be deleted in /rC/ 
clusters when in coda position followed by consonant-initial clitic, as in dármelo ‘give.me.it’ 
comprárselo ‘buy.yourself.it’; deletion of the rhotic before a definite article, as reported for 
Northern Peninsular Spanish and for Ecuadorian Spanish, has not been reported for Costa 
Rican Spanish.
Overall, in Spanish the tap has undergone various phonetic processes, resulting in 
the following phonetic realizations across various Spanish dialects and depending on specific 
phonetic contexts:
a. tap lateralization in coda position
b. vocalization of taps in pre-consonantal position and word-finally in words with final 
stress
c. insertion of a vowel-like segment between the consonant and the rhotic in /Cr/ onset 
clusters (a svarabhatki segment)
d. coarticulation in complex Cr clusters, mostly /t|/ and /d|/
e. insertion of an intervening svarabhakti vowel in /rC/ clusters in coda position
f. assibilation of taps before consonants in /rC/ cluster in coda position before consonants 
with the same voicing value
g. assibilation of syllable-final taps, specially phrase-finally
h. devoicing of word-final taps
i. deletion of taps in coda position before a definite article heading a following noun 
phrase
j. deletion of taps in coda position preceding a pronominal clitic pronoun that starts with 
a consonant
VÁSQUEZ: Children’s use of Spanish taps: A naturalistic study... 221
This tendency of the tap to undergo various assimilation processes in adult Spanish 
suggests that Spanish-speaking monolingual children might replicate similar patterns in their 
acquisition of the tap. Hence, it is crucial to review the literature on the acquisition of taps by 
Spanish-speaking monolingual children, in particular, that related to dialectal variation.
3.1. acquisition of spanish taps
Although few studies to date have reported on dialect-specific patterns with regards to 
the acquisition of Spanish rhotics (Ciccia-Gabillo, Montezuma del Castillo, Elías-Ulloa, Gabillo-
Ciccia, Echenique-Herrera, Raffo-Castro & Seminario-Olórtigue, 2006; Gómez- Fernández, 
2004), several studies have demonstrated that rhotics are late acquisition, though by age 4;0 
approximately, the majority of Spanish-speaking monolinguals reportedly produce the tap 
correctly (Acevedo, 1993; Bosch-Galcerán, 1983; Jiménez, 1987; Mason, Smith & Hinshaw, 1976).
Before full mastery of the tap is achieved, however, some authors identify a tendency 
to substitute it for a central lateral approximant [l], as liquids are normally acquired before 
rhotics (by Fernández-Aragón, Gutiérrez-Coto, Morgan-Mora, Romero-Zúñiga & Zadwaski-
Wisniewski, 1994; Gómez-Fernández, 2004; Moralejo, 2007; Yavas, 2004, in Proctor, 2009). 
Several authors additionally observe simplifications of the tap in consonant clusters (e.g., 
Anderson & Smith, 1987; Bosch-Galcerán, 1983; Fernández-Aragón et ál., 1994; Gildersleeve-
Newman, 2008; Goldstein & Iglesias, 1996; Gómez-Fernández, 2004). Fernández-Aragón et 
ál. report substitutions of the tap for [s], [n] and [l] in word-final contexts in the speech by a 
group of Costa Rican children ages 5;6 to 6;6, all diagnosed with speech problems. Syllable-
initially, they reported substitutions with [l] as well as frequent omissions (in 14% of the overall 
productions), whereas inter-vocalically, they also found substitutions with [l]. In /Cr/ clusters, 
Fernández-Aragón reported substitutions as well as omissions (e.g., grada à gada/glada ‘step’).
Regarding dialect-specific patterns, Ciccia-Gabillo et ál. (2006) describe what they 
refer to as semi-consonants ([semiconsonantación]) used to substitute for rhotics in child 
Peruvian Spanish; specifically, they report that instead of the trill and the tap, children 
normally produce either /j/ or /j^/. These researchers additionally describe the reduction of /tr/ 
clusters to a consonant followed by a vowel-like segment /tj/ in that Spanish dialect.
Similarly, Gómez-Fernández (2004) reports on naturalistic data by a group of children 
between ages 1 and 3 in Seville. He found that, between ages 1 and 1;5, the tap was omitted 
frequently, although sometimes it was substituted by [l], [D], and [B]; between ages 1;5 and 2, // 
was omitted in /Cr/ clusters as well as word-finally, whereas between ages 2;0 and 2;5, it was 
mostly correct, except for /Cr/ clusters as well as in final position where it tended to be omitted. 
Between ages 2;5 and 3, the tap was generally used correctly, although sporadic omissions and 
substitutions were found, mainly in /Cr/ clusters; such omissions were child-specific.
In sum, Spanish monolingual children initially omit taps; during a second stage of 
acquisition, taps are substituted by a central lateral approximant, which appears to be reported 
for all Spanish dialects, as well as by segments that appear to be language-specific as in /D/ 
and /B/ for Seville. Finally, taps tend to be omitted word-finally as well as in consonant clusters.
Given that the tap is a late acquisition phenomenon and that it evidences various 
assimilation processes across Spanish dialects, this study was designed to examine the use 
of the Spanish tap by a group of Costa Rican children ranging in age from 3;0 to 5;6 and to 
observe whether what has been reported in previous studies as well as dialect-specific patterns 
are evidenced.
Filología y Lingüística 40 (1): 217-228, 2014/ ISSN: 0377-628X222
4. the study
4.1. subjects and data collection
The data analyzed here come from audio-recordings of 15 children (6 girls and 9 boys) 
aged 3;0 to 5;6, as they engaged in naturalistic play sessions with either a researcher or with 
peers. The sessions ranged from 20 to 40 minutes and there was an interval of two months 
approximately between each audio-recording; in total, four sessions per child were used for 
the analysis. The children were all monolingual Spanish-speaking children from the Western 
region of Costa Rica attending either public day-care centers or public pre-schools in the 
Naranjo, Palmares, San Ramon, and Zarcero counties.
After transcribing each recorded session, all the words requiring taps (i.e., henceforth 
referred to as target words) were carefully transcribed by using the International Phonetic 
Alphabet, and the various patterns of substitution and omission were carefully recorded. The 
transcriptions were conducted by the author, a well-trained linguist, and they were checked 
for reliability by the researcher and her assistant. When in doubt regarding the pronunciation 
of a given word, both the researcher and her assistant listened to the recording and decided 
on the most accurate transcription to use. In some instances, due to external noise mainly, the 
recordings were not clear and hence, those short sections were not included in the analysis.
4.2. analysis and Results
A total of 1080 examples involving taps were analyzed, 80% (i.e., 874) of which were 
target-like (i.e., the tap was phonetically realized as []). In fact, three of the children used the 
tap correctly always.
The second most frequent pattern found consisted of omissions of the target sound; 
specifically, 14% (i.e., 151) of the total number of target words lacked the tap, as illustrated in 
examples 1 through 5 below.
Examples:
(1) colorin colorado à coloin coloao
(2) cortando à cotando ‘cutting’
(3) hormigas à homigas ‘ants’
(4) frijol à fijol ‘bean’
(5) negrita à neguita ‘black.diminutive’
By and large, these omissions occurred word-medially, as only 7 examples of word-
final tap omissions were found. Similarly, 8 types of substitutions were identified, many of 
which consisted of isolated or very few examples, as described below:
a. The tap was substituted by a central approximant [l] 7% of the time, as illustrated in 
examples 6 through 10.
Examples:
(6) grabar à glabal ‘record.infi.’
(7) jugar à jugal ‘play.inf.’
(8) ahora à ahola ‘now’
(9) cocodrilo à cocolilo ‘crocodile’
(10) comieron à comielon ‘eat.3rd.pl.past.nm
VÁSQUEZ: Children’s use of Spanish taps: A naturalistic study... 223
For some of these examples, as in 9, one could argue that it results from regressive 
assimilation, as the onset of the following syllable starts with a central lateral 
approximant [l].
b. [] was additionally substituted by [n], in a small number of the target words, namely 
14 (i.e., 1.29%); the examples were found in the speech of 3 children (ages 4 to 5;6), 
and they included: convierten à convienten (‘convert.3rd.pl.’), lastimar à lastiman 
(‘hurt.inf.’), hacer à hacen (‘do.inf.’), quebrar (queblan) (‘break.inf.’), and llorar à 
llolan (‘cry.inf.’). Only one of the examples consisted of a substitution word-medially, 
whereas the rest were all word-final.
c. [] was further substituted by a voiceless alveolar stop [t] by a single child at age 4;5. 
This pattern only represented 0.37% of the overall target words, as it was only found 
in 4 examples, namely mojar à mojat (‘wet.inf.’), llover à llovet (‘rain.inf.’), romper 
à llompet (‘break.inf.’), and resbalar à llesbalat (‘slip.inf.’). Interestingly, this 
substitution occurred phrase finally only.
d. [] was moreover substituted by [d] in the speech of a single child aged 4;10 in the 
utterance ir al patio à id al patio (‘go.inf. to the playground’); this substitution only 
constituted 0.09% of the total target words.
c. [] was also vocalized in 2 examples by a single child aged 4;3 to 5;1, namely in quebró 
à quebjó (‘break.1st.sg.past’) and bruja à bjuja (‘witch’).
e. [] was further substituted by a post alveolar affricate [t] in a single example by a child 
aged 3;4, namely in otro à ocho (‘other’).
f. [] was devoiced word-finally in 5 of the target words, produced by 3 of the children. 
More specifically, in the words jaguar à jaguar # and pasar à pasar #; and three 
examples with comer à comer #. Just as reported for adult Costa Rican Spanish, all the 
examples occurred phrase-finally.
g. [] was finally also substituted by a voiced interdental fricative [D]. This pattern was 
only found in 0.37% of the total target words, and it was found in four examples in the 
speech of a single child, namely ahora à ahoDa (‘now’), espere à espeDe (‘wait.1st.
sg’), número à númeDo (‘number’).
Overall then, one can state that although by and large the tap was used accurately, 
as evidenced in 80% of the total number of target words, the second most frequent pattern 
identified consisted of omissions. In turn, substitutions by a central lateral approximate 
constituted the third most common pattern spotted. Furthermore, the tap was substituted by 
seven different sounds, although these substitutions might not truly be ‘patterns’ per se, given 
that they only amounted to 1.85% altogether and were present in the speech of a few of the 
children. Of these substitutions, at least [n] and [n] were only found in final codas, the same 
as the assibilated voiceless rhotic.
On the other hand, the data revealed a curious context where taps were typically 
omitted, and which was not counted as part of the omissions reported above; they were 
analyzed separately due to their particular nature. These omissions occurred in word-final 
Filología y Lingüística 40 (1): 217-228, 2014/ ISSN: 0377-628X224
codas in infinitive verbs before a consonant-initial enclitic (i.e., clitic pronoun which attaches to 
the end of verbs). Of the overall 28 examples of words requiring a tap in such a context (2.59% 
of the total target words), 26 were missing the tap. The question then rose as to whether this 
pattern was related to the children’s acquisition of clitc pronouns1. Hence, all examples of verb 
forms containing enclitics were analyzed.
Interestingly, upon examining these children’s use of enclitics, it was evident that they 
were using verb constructions that included not only one but sometimes two enclitics. In the total 
14 examples identified and which did not involve a final tap, no errors regarding omissions or 
gender and number agreement were identified, as shown in the examples in 11 through 16 below:
(11) cuéntenme ‘tell.2nd.pl.1st.sg.’
(12) póngase.los ‘put.2nd.sg.refelxive.3rd.sg.masc.’
(13) llevándose.lo ‘bring.prog.refelxive.3rd.sg.masc.’
(14) poniéndose.los ‘put.prog.refelxive.3rd.pl.masc.’
(15) comiéndose.lo ‘eat.prog.refelxive.sg.masc.’
(16) siéntese ‘sit.imp.2nd.sg.
Despite the evidence that these children were able to produce verb phrases with 
enclitics without problems, a significant number of examples where children omitted, not 
the enclitic but the tap in infinitive verb forms when preceded by an enclitic starting with a 
consonant sound were found; these are listed below:
ensenar.los à enseñalos
(‘teach.inf.them.they.mac.’; 2 examples)
jugar.lo à jugalo
(‘play.inf.it.masc.’)
oir.los à oilos
(‘hear.inf.them.masc.’)
comer.los à comelos
(‘eat.inf.them.masc.’)
comer.se.la à comese.la 
(eat.inf.reflexive.it.fem.)
comprar.lo à compralo
(‘buy.inf.it.masc.’)
contar.le à contale
(‘tell.inf.3rd.sg.’)
comer.la à comela
(‘eat.inf.it.fem.’)
mojár.se.lo à mojase.lo
(‘getwet.inf.refelxive.it.masc.’)
decir.les à deciles
(‘tell.inf.them)
escribir.le à escribile
(‘write.inf.3rd.sg.’)
bajar.la à bajala
(‘takedown.inf.it.fem.’)
preguntar.le à preguntale
(‘ask.inf.it)
montar.se à montase
(‘ride.inf.refelxive’)
peinar.se à peinase
(‘comb.inf.refelxive’)
ir.me à ime
(‘go.inf.me’)
rajar.la à rajala
(‘tear.inf.it.fem.’)
caer.se à caese
(‘fall.inf.reflexive’)
comer.los à comelos
(‘eat.inf.them.mac.’)
comér.se.lo à comese.lo
(‘eat.inf.refelxive.it.masc.’)
cortar.le à cortile
(‘cut.inf.3rd.sg.’)
pegar.le à pegale
(‘hit.inf.3rd.sg.’)
poner.me à poneme
(‘put.inf.myself’)
contár.se.lo à contuse.lo
(‘cut.inf.refelxive.it.masc.’)
One possible account for these omissions might be the tendency in early child speech to 
simplify syllables to the basic syllable structure, CV, and to reduce the total number of syllables, 
often resulting in final consonant deletion, as stated in Bosch-Galcerán (1983). Nonetheless, 
upon examining all the words containing a tap followed by another consonant in the speech of 
the children who omitted the tap before an enclitic, as in ver.de ‘green’ and par.te ‘part’, it was 
found that those taps were not omitted despite also constituting complex CVC syllables.
A final possibility entertained was that such omissions were only used in rapid speech, 
but that the tap would surface if the children produced these infinitive verb + enclitic forms 
VÁSQUEZ: Children’s use of Spanish taps: A naturalistic study... 225
slowly and in a paused fashion. This, however, was not the case, as whenever the researcher 
heard one of these omissions, she prompted the children to slowly repeat the constructions; 
they would then repeat the structure in slow motion, but always leaving out the infinitival tap, 
as shown in the excerpt below:
*CHI: es que quiero pa’ contáselo. ‘I want it to sing it to you’
*R1: cómo? Diga: para con tár-se-lo (sounding out each separate syllable) ‘what, say: 
to tell it to you’
*CHI: para con tá se lo (separating each syllable through a pause)‘to sing it to you’)
Recall that these omissions have been reported in adult Spanish, and hence it is 
possible that these children were simply reproducing what they heard in their input, but I would 
like to entertain the hypothesis that these omissions are directly related to the complex nature 
of enclitic pronouns, given that in comparable syllabic contexts (CVC), omission of taps was 
not found.
5. Conclusions
This study was designed to observe the use of Spanish taps by a group of Costa 
Rican Spanish-speaking children ages 3;0 to 5;6. The study revealed that the majority of 
these children had acquired the tap, though common omissions as well as a series of mostly 
unsystematic substitutions were identified.
Most of the tendencies found in these data have been reported in previous studies. For 
example, omissions of taps are common in child Spanish, as evidenced in Fernández-Aragón 
et ál. (1994), Moralejo, (2007), and Yavas (2004) in Proctor (2009). The same is true for 
substitutions of the tap for a central lateral approximant, which is normally acquired before 
other approximants, including taps and trills (e.g., Fernández-Aragón et ál., 1994; Gómez-
Fernández, 2004; Moralejo, 2007; Yavas, 2004, in Proctor, 2009).
With regards to language-specific trends, substitutions of the tap for [n] were 
reported for Costa Rican Spanish-speaking children with speech difficulties in Fernández-
Aragón et ál. (1994), whereas vocalizations of the tap were reported for Peruvian Spanish 
in Ciccia-Gabillo et ál. (2006). Interestingly, substitutions with [D] have only been reported 
for Sicilian Spanish-speaking children in Gómez-Fernández (2004); the voiced interdental 
fricative phoneme is part of that Spanish dialect and hence one could argue that it results from 
the children’s imitation of their input. In this study, which included Costa Rican children, 
nonetheless, [D] is not found in the input.
Nonetheless, the substitutions for a voiceless assibilated rhotic in final coda position 
could be a reflection of these children’s input, as this tendency is found in adult Costa Rican 
Spanish (Calvo-Shadid & Portilla-Chaves, 1998; Vásquez-Carranza, 2006). The same is true for 
the deletion of taps in infinitive verb forms before nominal enclitics, as this pattern has been 
reported for Costa Rican Spanish Vásquez-Carranza (2006); yet, a different account which has 
been entertained here is that, these deletions result from something inherit in nominal enclitics, 
as taps in similar phonetic contexts do not get deleted consistently as do these taps (recall that 
although no previous studies have reported this tendency in child speech, some studies describe 
it in other adult Spanish dialects, namely Penny, 1969 [cited in Bradley, 2005] and Bradley, 1998). 
This possibility, nonetheless, has yet to be explored and explained in future research.
Filología y Lingüística 40 (1): 217-228, 2014/ ISSN: 0377-628X226
Overall, this study confirms findings in previous reports that the tap is a late 
acquisition phenomenon and that, just as is the case with the acquisition of trills, taps are often 
substituted by approximations which involve less articulatory effort, and some of which derive 
from the children’s input that itself contains assimilations.
notes
1. Clitics are described as a set of features which denote number, gender, person, and case, and Spanish 
has a large set of clitics, namely, a) verb objects (direct objects, as in lo tengo ‘it I’ve.got’ and indirect 
objects, as in le digo que venga ‘him/her I tell to come’) and non-objects (reflexive se, as in se lava 
‘itself washes’, reciprocal se, as in se quieren ‘each other they love’, impersonal se, as in se come bien 
‘you.impersonal eat well’, middle-passive sentences se, as in se venden estas casas ‘are for sale, these 
houses’, and lexical aspect se, as in se comió la sopa ‘he/she/it has already eaten the soup’ (Examples 
taken from  Gutiérrez-Clellen & Simon-Cerejido, 2007, pp. 319, 320)
Clitics are also referred to as weak pronouns, and they are described in Fujino & Sano (2002) as 
“unstressed elements that attach to a host verb and serve a pronominal function” (p. 71); this is true at 
least for enclitics (i.e., clitic pronouns which attach to the end of a verb).
Regarding the acquisition of clitics, Müller & Hulk (2001) and Simon-Cerejido & Gutiérrez-Clellen 
(2007), report that their use in child speech may vary depending on the amount of exposure, as clitics 
are optional in some language dialects and children might hence not evidence them in their speech.
In fact, Grinstead (2000), Domínguez (2003) and López-Ornant (1994) observe that clitics are used 
correctly from the beginning, whereas Fujino and Sano (2002) report that clitics are initially absent or 
infrequent in early child Spanish, but that at a certain point, there is a clear increase in children’s use 
of clitics and they become productive. Grinstead further adds that Spanish monolingual children do 
not commit errors of commission (exchanging an accusative clitic for a dative clitic, for example) and 
rarely evidence omission errors. He additionally points out that in early Spanish, imperative verb forms 
occur to the left of clitics (e.g., dame ‘give.me’), whereas finite verbs occur to the right of clitics (e.g., la 
pone aquí ‘it 3rd.sg.pron. put here’). Torrens & Wexler (1996) point out that children rarely make clitic 
placement errors with both finite and non-finite verbs. Finally, Lyczskowski (1999, cited in Torrens & 
Wexler, 1996) argue that malformed or misplaced clitics are rare in child Spanish.
References
Acevedo, M.A. (1993). Development of Spanish Consonants in Preschool Children. Journal of 
Childhood Communication Disorders. 15 (2), 9-15.
Anderson, R.T. & Smith, B.L. (1987). Phonological development of two-year old monolingual 
Puerto Rican Spanish-speaking children. Journal of Child Language. 14 (1), 57-78.
Bosh-Galcerán, L. (1983). El Desarrollo Fonológico Infantil: Una prueba para su evaluación. 
Anuario de psicología. 28 (1), 85-114.
Bradley, T.G. (1998). Assibilation in Ecuadorian Spanish. In J.M. Authier, B.E. Bullock & L.A. 
Reed (Eds.). Formal Perspectives on Romance Linguistics. Selected papers from the 
28th Linguistic Symposium on Romance Languages. (57-71). University Park.
Bradley, T.G. (2001). A typology of rhotic duration contrast and neutralization. By M.J. Kim 
and U. Strauss (Eds.) Proceedings of the 31st North East Linguistics Society. (79-97). 
Amherst, MA: GSLA Publications.
Bradley, T.G. (2005). Systemic markedness and phonetic detail in phonology. In R. Gess & E. 
Rubin (Eds.). Experimental and theoretical approaches to Romance linguistics. (41-
62). Amsterdam: John Benjamins.
VÁSQUEZ: Children’s use of Spanish taps: A naturalistic study... 227
Bradley, T.G. & Schmeiser, B. (2003). On the phonetic reality of Spanish // in complex onsets. 
University of California: Working draft.
Calvo-Shadid, A. & Portilla-Chaves, M. (1998). Variantes retroflejas de // y /r/ en el habla 
culta de San José. Revista Káñina. 22 (1), 81-86.
Ciccia-Gabillo, C., Montezuma del Castillo, O., Elías-Ulloa, J., Gabillo-Ciccia, G., Echenique-
Herrera, M., Raffo-Castro, G. & Seminario-Otórtoga, A. (2006). Las divergencias 
fonológicas entre el habla adulta el de los niños en el español hablado en Lima (Perú). 
Revista de Logopedia, Foniatría y Audiología. 26 (4), 215-230.
Domínguez, L. (2003). Interpreting reference in early acquisition of Spanish clitics. In S. 
Montrul & F. Ordóñez (Eds.). Linguistic Theory and Language Development in 
Hispanic Languages, Papers from the 5th Hispanic Linguistics Symposium and the 
4th Conference on the Acquisition of Spanish and Portuguese. (212-228). University 
of Illinois.
Fernández-Aragón, A., Gutiérrez-Coto, L., Morgan-Mora, G., Romero-Zúñiga, I.L. & Zawaski-
Wisniewiski, S. (1994). Problemas en la la adquisición de la lengua maternal: 
fonética y fonología. (Tesis de Licenciatura sin publicar). Universidad de Costa Rica.
Fujino, H., & Sano, T. (2002). Aspects of the null object phenomenon in child Spanish. In A.T. 
Pérez-Leroux & J.M. Liceras (Eds.). The acquisition of Spanish Morphosyntax. (67-
88). Dordrecht: Kluwer Academic Publishers.
Gildersleeve-Newman, C.E., Kester, E.S, Davis, B.L. & Peña, E.D. (2008). English speech 
sound development in preschool-aged children from bilingual English-Spanish 
environments. Language, Speech, and Hearing Services in schools. 39 (3), 314-328.
Goldstein, B & Iglesias, A. (1996). Phonological Patterns in Puerto Rican Spanish-speaking 
children with phonological disorders. Journal of Communication Disorders. 29 (5), 
367-387.
Gómez-Fernández, D. (2004). La adquisición de las líquidas en los niños de Sevilla y su 
provincia: Desde los doce a los treinta y seis meses. Revista de Filología y su 
Didáctica. (27), 125-185.
Grinstead, J. (2004). Subjects and interface delay in child Spanish and Catalan. Language. 80 
(1), 40-72.
Gutiérrez-Clellen, V.F. & Simon-Cerejido, G. (2007). The discriminant accuracy of a 
grammatical measure with Latino English-speaking children. Journal of  Speech, 
Language, and  Hearing Research. 50 (4), 968-981.
Harris, J. (1969). Spanish Phonology. Cambrigde: Massachusetts.
Jiménez, B.C. (1987). Acquisition of Spanish consonants in children aged 3-5 years, 7 months. 
Language, Speech, and Hearing Services in Schools. 18, 357-363.
Ladefoged, P. (1971). Preliminaries to Linguistic Phonetics. London & Chicago: University of 
Chicago Press.
Ladefoged, P. (2001). Vowels and Consonants: An introduction to the sounds of languages. 
Oxford: Blackwell Publishers.
Ladefoged, P. & Maddieson, I. (1996). The Sounds of the World’s Languages. Oxford: Blackwell.
Filología y Lingüística 40 (1): 217-228, 2014/ ISSN: 0377-628X228
López-Morales, H. (1992). La investigación dialectal sincrónica en Hispanoamérica: presente y 
futuro. Actas del Congreso de Lengua Española. Sevilla, Instituto Cervantes. (767-787).
López-Ornat, S. (1994). La Adquisición de la Lengua Española. Madrid: Siglo XXI.
Mason, M., Smith, M. & Hinshaw, M. (1976). Medida Española de articulación [Measurement 
of Spanish Articulation]. San Ysidro, CA: San Ysidro School District. 
Moralejo, R. (2007). Adquisiciones fonético-fonológicas en niños de 4 años de edad. http://
www.espaciologopedico.com/ [Available June, 2010].
Müller. N. & Hulk, A. (2001). Crosslinguistic Influence in Bilingual Language Acquisition: 
Italian and French as Recipient Languages. Bilingualism: Language and Cognition. 
4 (1), 1-21.
Proctor, M.I. (2009). Gestural characterization of a phonological class: The liquids. 
(Unpublished Doctoral Dissertation). Yale University.
Quilis, A. (1999). Tratado de Fonología y Fonética Españolas. Madrid: Editorial Gredos.
Recasens, D. (1991). On the production characteristics of apicoalveolar taps and trills. Journal 
of Phonetics. 19, 267-280.
Sánchez-Corrales, V. (1986). Escisión fonológica de /®/ en el español de Costa Rica. Revista de 
Filología y Lingüística. 12 (2), 129-133.
Stockwell, R.P. & Bowen, J.D. (1965). The Sounds of English and Spanish. London & Chicago: 
University of Chicago Press.
Torrens, V. & Wexler, K. (1996). Clitic Doubling in early Spanish. In Proceedings of 20th 
the Boston University Conference on Child Language Development. (780-791). 
Sommerville MA: Cascadilla Press.
Umaña-Aguilar, J. (1981). Variable vibrants in middle-class Costa Rican Spanish. (Unpublished 
Masters Thesis). Georgetown University.
Vásquez-Carranza, L.M. (2006). On the phonetic realization and distribution of Costa Rican 
rhotics. Revista de Filología y Lingüística. 32 (2), 231-309.
Willis, E.W. 2006. Trill Variation in Dominican Spanish: An Acoustic Examination and 
Comparative Analysis. By N. Sagarra & A.J. Toribio (Eds.) Selected Proceedings 
of the 9th Hispanic Linguistics Symposium. (121-131). Sommerville: Cascadilla 
Proceedings Project.