clac 12/ 2002


David Eddington


University of New Mexico


            The VOS form of Classical Latin is manifest in Contemporary Spanish in two distinct ways.  Vos in Spanish emerged from Latin VOS as a second person singular manner of address, and differed from tú in that the latter was more informal and employed with inferiors.  In the seventeenth, and possibly as late as the eighteenth century, the use of vos disappeared from many parts of the Spanish speaking world (Lapesa 1984:579), although it is still widely used in many areas of South and Central America (Rona 1967). Vosotros retained the plurality of Latin VOS to become an informal manner of address.  However, in contemporary Peninsular Spanish it has lost a lot of ground to the ustedes  form.  The only region to conserve the vosotros form is Spain, yet even within Spain it is not universal; the western half of Andalusia has lost it in favor of ustedes (Lapesa 1984:512).  The imperative of Latin VOS was marked with the verb-final morpheme -TE (AUDITE 'hear').  This morpheme, with the voicing of intervocalic /t/ and deletion of final /-e/ common in the evolution of Spanish, became the final /-d/ morpheme indicative of the vosotros imperative (comed, bebed).


             The objective of this paper is to examine other varieties of this imperative, namely the vos imperative, and the final /-r/ vosotros imperative, and to give a morphophonemic explanation of why these varieties have arisen in contrast to the final /-d/ variety of the standardized language.  For clarity's sake I will define each of the imperative varieties I will deal with.  What will be referred to as the -d variety imperative is the variety that is prescribed by traditional grammars.  In this variety a /-d/ replaces the /-r/ of the infinitive in the following manner:

(1)       a. decir > decid

                                    b. venir > venid


With reflexive verbs, however, the -d is omitted, most likely to avoid confusion with the past participle:


                        (2)       a. acercarse > acercaos

                                    b. divertirse > divertíos

            The vos imperative will be called the null variety since no word final morpheme appears after the theme vowel:

                        (3)       a. decir > decí

                                    b. venir > vení


Vos imperatives employ the reflexive pronoun te in place of os:


                        (4)       a. acercarse > acercate

                                    b. divertirse > divertite


The final -r variety does not differ from the infinitive form of the verb, and the clitic pronoun os is added to the verb:


                        (5)       acercarse > acercaros


Most of the recognition that the -r variety imperative has recieved has been prescriptive in nature; it is recognized by the Royal Spanish Academy only as a vulgar form of speech since it is not found in literary works except as an imitation of uncultured speech (Real Academia 1985:460).  However, indications are that the -r variety is not uncommon in the peninsula.  It has been documented in Aragón, Navarra, La Rioja and Andalucía (Alvar, et al.1980: map 1837; Alvar et al. 1973: map  1.719), and I have observed it in varying degrees in the speech of Spaniards from many regions, and social classes.


            I submit that the existence of the -r and null variety imperatives derived from Latin -TE can

be explained by considering the frequency of the consonant clusters that arise when the clitic pronouns are used in conjunction with the imperative.  The consonant clusters formed by the combination of the final -d of the vosotros imperative and the initial consonants of the clitic pronouns are highly anomalous in the Spanish lexicon, while consonant clusters formed between the clitic pronouns and the final -r of the -r variety, as well as those formed with the null variety imperative, are common clusters.


            Consider the consonant clusters that are formed with the final -d of the -d variety imperative combined with the initial consonants of the clitic pronouns, (n-, s-, l-, m-, t-).  The clusters -dn-, -ds-, -dl-, -dm-, and -dt are the possible combinations (i.e. dejádnoslo, dádselo, bajadlo, ayudadme).  Of the above clusters, -dt- may be excluded on pragmatic grounds since it is not a viable option (i. e. *ponedte).  When a Spanish speaker encounters one of the clusters from the -d variety imperative, s/he is faced with a cluster which is highly anomalous elsewhere in the language.  These anomalous clusters, -dn-, -ds-, -dl-, and -dm-, occur almost exclusively when the -d variety imperative is combined with a clitic pronoun.  The -r variety imperative, however, yields the more common clusters -rn-, -rs-, -rl-, and -rm-.  Along the same lines, the null variety imperative also yields the more common clusters -Vn-, -Vs-, -Vl-, and -Vm-.


            In order to determine whether the clusters -dn-, -ds-, -dl-, and -dm- were as uncommon as they appear I searched for words that contained them in the Alameda and Cuetos (1995) frequency dictionary.  It was found that outside of -d imperatives, the cluster -dm- is found in cadmio and the English loan word bádminton.  Other that those cases, only a few other words, all stemming from the prefix ad-, were found: admitir, administrar, admonestación, admirar, admonículo (and other words with the same roots).  The rarity of the -dm- cluster is further demonstrated by the fact that the /d/ in the above words is often deleted in popular speech:


                        (6)       a. admitir [amitír]

                                    b. admonestación [amonestaθjón]


I argue that deletion is a way of bringing the odd cluster into line with more natural clusters of the language.  Much more common are words with the cluster -rm- that occur in the -r variety imperative: ermitaño, hormiga, hormona, hermosura, armada.


            Outside of the vosotros imperative, I found no words that contained the cluster -dn-.  That -dn- is an anomaly in Spanish is also attested to historically.  Occasionally apocope of atonic vowels in Latin should have yielded the cluster -dn- in Spanish, however, -dn- was avoided by metathesis into -nd-:


                        (7)       a. candado, *cadnado < CATENATU

                                    b. lindo, *lidmo < LEGITIMU

                                    c. rienda, *riedna < *RETINA

                                    d. serondo/seroño, *serodno < SEROTINU (Lloyd 205)


On the other hand, words that contain the cluster -rn- of the -r variety imperative are abundant: carne, infierno, ternura, etc.

            The cluster -ds- is uncommon as is seen by the fact that so few words contain it.  Only the words ascribir, adsorber, and adstrato and their morphemic relatives were found in the frequency dictionary to contain the cluster -ds-.  Once again, -rs- is common where -ds- is not: curso, reverso, dorsal, persona, etc.   In the frequency dictionary, adláteres is the only non-imperative containing the cluster -dl-.  Examples of the aversion to this cluster can also be found historically.  In Golden Age Spanish, the anomalous cluster -dl- was eliminated by metathesis of the elements of the cluster (Menéndez-Pidal 301):


                        (8)       a. dadle > dalde

                                    b. ponedlo > poneldo


Lloyd notes that this was indeed done to eliminate a low frequency consonant cluster (1987: 360). In like manner, this cluster suffered metathesis as Latin words developed into Spanish.  When voicing of /t/ and loss of unstressed vowels would have normally resulted in the formation of the cluster -dl-, methathesis of the elements of this cluster yielded instead -ld-:


                        (9)       a. CAPIT(U)LU > cabildo, *cabidlo


                                    b. SPAT(U)LA > espalda, *espadla


                                    c. TIT(U)LARE > tildar, *tidlar (Lathrop 136)



            Thus far, I have demonstrated that when a Spanish speaker uses the -r variety vosotros imperative instead of the -d variety, it is done in order to avoid consonant clusters that are very unusual in the language.  The use of the null variety in place of the -d variety similarly serves to eliminate these unusual clusters.  That they are unusual is evidenced by their low frequency of occurrence within words. That they are treated as unusual clusters when they fall between imperatives and clitic pronouns is evidenced by the existence of the -r and null variety imperatives which do not contain them.  Metathesis of the clusters -dl-, and -dn- as well as the frequent loss of /d/ in words containing the cluster -dm- is further evidence that these clusters are odd.


            The crux of my argument rests on the assumption that the interplay between the final -d of the imperative and the initial consonants of the clitic pronouns is significant enough to allow the emergence of alternative imperative forms.  This exact sort of interplay has been thought to be responsible for the phonetic evolution of the clitic pronoun os < VOS.  The first instances of vos losing its initial consonant to become modern os occurred immediately following the final -d of the vosotros imperative (Lathrop 1984: 155).  The resulting cluster, -dv-, was anomalous enough to pressure /b/ to drop resulting in a much more normal consonant cluster.  From this post-imperative position, os < vos was then generalized to all other positions in the language.


            Besides the uncommon consonant clusters discussed, the -d variety imperative is subject to another force which could have led to the appearance of alternative imperative forms.  In syllable final position [-ð] is subject to various modifications, and in -d variety imperatives, /d/ always occurs in this weak position.  In popular speech /d/ is often devoiced in syllable final position, and when followed by a voiceless consonant (Alarcos Llorach 184):


                        (10)     a. verdad [berðáθ]

                                    b. ataúd [ataúθ]


Many speakers omit /d/ when it appears in final absolute position, and when it is followed by a voiceless consonant: (Alarcos Llorach 1974:185)


                        (11)     a. Madrid > [maðrí]

                                    b. hermandad > [ermandá]


Moreover, in popular and unmonitored speech word final [-ð] will be dropped even before voiced consonants:


                        (12)     a. un ataud blanco [unataúßláŋko]

                                    b. la edad granuja [laeðáγranúxa]


            It is not surprising to see final -d become modified in final position given that the only consonants that may occupy word final position in Spanish are /l, r, n, s, ?, d/, and in this position /d/ is realized as a fricative [ð], which makes it anomalous in that it is the only voiced fricative that may occur in word final position.  The widespread tendency for final [ð] to be deleted or modified in word final position must certainly have had some bearing on the development of an imperative that lacks final [ð] such as the null variety imperative of contemporary voseo.


            Another possible factor that should be mentioned to round out this discussion is that of analogy.  Spanish allows the infinitive of the verb to be used as an imperative when the imperative is directed to an unknown party.  Such unmarked imperatives are often found in the directions for household items, recipes, and other commands directed at the general public (e.g. tirar de la cadena, verter medio litro de agua).  Given the problems associated with the -d variety, it is likely that the -r variety developed as a result of the analogical force exerted upon it by the unmarked infinitival imperative.


            The -r and null variety imperatives in Spanish should not simply be attributed to careless speech habits of the masses, but rather, to linguistic forces arising from the structure of the language itself.  The phoneme /d/ is unstable in final position, and this comes to a head in the -d variety imperative.  Final /-d/ is subject to devoicing and deletion, and in the case of voseo, this deletion became generalized as the phonetically null imperative morpheme.  Consonant clusters formed with the final /-d/ of the -d variety imperative and the initial consonants of the clitic pronouns are highly anomalous in Spanish, while clusters formed under the same conditions with the -r variety are more common.  Finally, the widespread use of the infinitive as an unmarked imperative may have been adopted as an alternative vosotros command form in order to avoid the problems associated with the -d variety imperative.



Works Cited

Alameda, José Ramón, and Fernando Cuetos. Diccionario de frecuencias de las unidades lingüísticas del castellano. Oviedo, Spain: University of Oviedo Press, 1995.

Alarcos Llorach, Emilio. Fonología española. 4th ed. Madrid: Gredos, 1974

Alvar, Manuel, A. Llorente, T. Buesa, and Elena Alvar. Atlas lingúístico y etnográfico de Aragón, Navarra y Rioja. Vol 12. Paracuellos del Jamara, Madrid: Departamento de Geografía y Lingüística, Institución Fernando el Católico de la Excma. Diputación Provincial de Zaragoza, 1980.

Alvar, Manuel, A. Llorente, and G. Salvador. Atlas lingüístico y etnográfico de Andalucía. Vol. 6. Granada: U. de Granada, Consejo Superior de Investigaciones Científicas, 1973.

Lapesa, Rafael. Historia de la lengua española. 9th ed. Madrid: Gredos, 1984.

Lathrop, Thomas A. Curso de gramática histórica española. Juan Gutiérrez and Ana Blas, trans. Barcelona: Editorial Ariel, 1984.

Lloyd, Paul M. From Latin to Spanish. Philadelphia: The American Philosophical Society, 1987 [Spanish trans. and notes Adelino Alvarez, Madrid 1993, Gredos].

Menéndez-Pidal, R. Manual de gramática histórica española. 18th Ed. Madrid: Espasa-Calpe, 1985.

Real Academia Española. Esbozo de una nueva gramática de la lengua española. Madrid: Espasa-Calpe, 1985.

Rona, José Pedro. Geografía y morfología del "voseo". Pôrto Alegre: Pontíficia Universidade Católica do Rio Grande do Sul, 1967.



© David Eddington. Circle of Linguistics Applied to Communication/ Círculo de Lingüística Aplicada a la Comunicación 12, November 2002. ISSN 1576-4737. Published in Anuario de Estudios Filológicos 14, 1991, 125-130 (A morphophonemic explanation of the ‘vos’ and final -r variety ‘vosotros’ imperative).


Contents of clac 12