The Spanish imperatives bendice and maldice of bendecir and maldecir are regular, that is, they are formed like all other regular singular informal imperatives, with a root + a theme vowel, and no ending. However, they are anomalous within their respective paradigms, since they do not align morphologically with the irregular forms of Sp. decir , e.g., pres. ind. digo , bendigo , maldigo , pret. dije, bendije, maldije, etc. Thus, since di is the irregular imperative of decir , one would expect irregular imperatives such as * bendí y * maldí from bendecir and maldecir . The matter is further complicated by the fact that bendi and maldi appear in texts of Alfonso X, el Sabio, from the late thirteenth century, which points to the fact that they were clearly the preferred variants of King Alfonso, and that they continue to this day in Galician and Asturian. A complete historical and philological analysis undertaken here, which takes into account evidence from non-standard varieties of Latin and Romance, together with processes of historical morphology such as syncretism and anti-homophony, explains the curiously regular Modern Spanish imperatives bendice and maldice .

Alongside the standard irregular preterite of Sp. andar (anduve 'I walked') there exist non-standard regular forms (and & eacute; 'I walked'). Scholars have assumed that the non-standard regular forms have resulted from analogy with the preterite of regular -ar verbs, agree that the irregular preterite of andar was formed by analogy with the etymological irregular preterite of estar, and have assumed that the earliest forms were the irregular ones. But if the preterite of andar was formed by analogy with that of estar, what were the etymological forms of the preterite of andar inherited from Latin? This question has not been raised hitherto because of "standard language bias", a phenomenon recently introduced by Maiden (2022). It will be revealed here that non-standard forms in and- appear by the early thirteenth century. Their early appearance, together with comparative Romance evidence, points to the likelihood that they were not formed by later processes of analogy and regularization, but that they were, and are, direct descendants of one of a number of possible Latin etyma. Finally, it will be explained why forms in anduv- eclipsed the original forms in and-, eventually becoming the standard, as well as why they have persisted as such to date.

Las formas del futuro (y condicional) de bendecir y maldecir son anómalas porque son regulares, por ejemplo bendeciré, maldeciré, mientras que en la mayoría de los demás tiempos y modos son irregulares y se alinean morfológicamente con las de decir, por ejemplo digo , ben digo , mal digo ; dije , ben dije , mal dije , etc. La regularidad del futuro / condicional de estos dos verbos se vuelve aún más sorprendente cuando uno considera que las formas del futuro / condicional eran irregulares en español antiguo, por ejemplo ben dir é, mal dir é. El presente artículo intenta explicar por qué estas formas se regularizaron en el futuro / condicional mientras que las formas de la mayoría de los demás tiempos y modos han permanecido irregulares y alineadas con las de decir. Se propone aquí que la regularización del futuro / condicional de bendecir y maldecir se debe a la estructura de sus participios pasados, dada la estrecha conexión entre el infinitivo y el participio pasado a lo largo de la historia de las lenguas romances, junto con el hecho de que, tanto el infinitivo como el participio pasado han estado unidos estructuralmente por el verbo auxiliar haber a través de la historia del español.

Abstract It is well known that the Spanish third person plural of -er and -ir verbs has two different preterite endings: primary -ieron, and the variant -eron, which appears after stem-final -j, e.g., dijeron, trajeron, tradujeron, etc., and that -eron arose from -ieron through absorption of the semiconsonant [j] into the Old Spanish stem-final palatal -x [ʃ], e.g., dixieron > dixeron. What is not so well known is that OSp. dixiemos, dixiestes, like dixieron > dixeron, likewise reduced to dixemos, dixestes, but unlike dixeron, did not survive. It is also to date unknown that the reduction of -ieron > -eron occurred in other grammatical categories, e.g., OSp. cogieron > cogeron, before reverting to -ieron. The present study will show when -eron first arose and became the predominant variant of the two after -x, the extent to which it arose in other grammatical categories, and will explain why vertical vocalic symmetry between OSp. dixiemos, dixiestes, dixieron, later dixemos, dixestes, dixeron, was ulimately broken, and how the loss of dixiemos/dixemos, dixiestes/dixestes on the one hand, and the triumph of dixeron on the other, are completely interconnected, ironically tied to the same historical linguistic phenomenon, and actually two sides of the same coin.

A lo largo de la historia del español, varios sustantivos han vacilado entre un género y otro, por ejemplo, amor, honor, calor, y color, mientras que otros han cambiado de género por completo, por ejemplo, labor y sal cambiaron de masculino a femenino, mientras que árbol, origen y valle cambiaron de femenino a masculino, y arte cambió de femenino a masculino solo en singular. El presente artículo tratará el sustantivo español árbol y: (1) mostrará el grado en que árbol apareció como masculino y femenino tanto en singular como en plural en español antiguo y cuándo el cambio a masculino fue completo tanto en singular como en plural; (2) considerará las diversas hipótesis que se han propuesto para explicar este cambio; y (3) analizará las formas del Lat. ARBO˘R en algunas lenguas románicas dentro de un contexto morfosintáctico y revelará cómo la estructura de estos sintagmas explica el cambio o mantenimiento de género en cada caso.

Abstract The origin of –y of Sp. soy, doy, voy, estoy remains unresolved to date. The present study reconsiders and elaborates the ‘post-verbal yo hypothesis’, originally proposed by Ford (1911), which has been rejected by most scholars because they have misunderstood it to suggest that the post-verbal subject pronoun yo became atonic and agglutinated to the verb, subsequently reducing to –y. It is suggested here that the phenomenon of ‘leftward palatal spreading’ from the post-verbal subject pronoun yo, with no agglutination nor reduction in form, first produced phonetic variants, e.g., so yo > ['soi̯-'yo], which were later morphologized and raised to allomorphic status before becoming full-fledged morphological variants upon extraction from the new syntagms, e.g., soy yo > soy (yo) > soy.

Abstract The pres. ind. paradigm of Sp. caber ‘to fit’ exhibits a synchronically irregular form in the 1st pers. sg., i.e., quepo, instead of a synchronically regular form derived from the infinitive, i.e., caber → *cabo. However, quepo is not considered at all historically irregular. Since the first historical grammar of Spanish, quepo has been understood to be a direct continuation of Lat. capiō, which apparently evolved through regular phonetic development, like pres. subj. capiam > quepa, sapiam > sepa. Nonetheless, one may question why quepo has not been replaced by *cabo in Modern Spanish given its extremely low frequency of occurrence, as forms of a language that occur infrequently are often regularized. A historical look at quepo reveals the following surprising facts: (1) Although pres. subj. quepa is attested from the earliest Old Spanish texts onward, quepo is absent from the written record throughout the Old and Medieval Spanish periods and does not appear until the end of the sixteenth century; (2) Regularized cabo served as the first person singular of the present indicative until then. The present study attempts to explain through well-established processes of historical morphology the late appearance of quepo and its continued existence in Modern Spanish.

Modern Spanish has twelve verbs that are considered synchronically irregular in the future indicative (and conditional) because they exhibit roots that are not comprised of the full form of the infinite (plus the verbal ending), but rather, syncopated forms such as habra (of haber), sabra (of saber), and other syncopated forms with modification of their resultant consonant sequences such as vendra (of venir), pondra (of poner), etc. In Old Spanish, virtually any second or third conjugation verb could undergo syncope and subsequent consonantal modification (if necessary), but only twelve remain. Conventional wisdom has held that only the most frequent of the Old Spanish irregular verbs have continued into Modern Spanish. However, when one considers the frequency of some of the twelve irregular future (and conditional) verbs of Modern Spanish, it becomes clear that high frequency of occurrence cannot be the only factor that has preserved all twelve of these verbs. The present study reveals a morphological factor that can account for the retention of the lowest frequency verbs of the twelve.

The Spanish noun arte is unique in that it is, for all intents and purposes, masculine in the singular, but feminine in the plural, e.g. el arte culinario vs las artes culinarias. This noun was feminine in Latin, and in Old Spanish as well, but because it has been preceded by the ?feminine el? (i.e., allomorph of the feminine definite article la) and has exhibited the ambiguous gender marker ?e, el arte (feminine) was eventually, and understandably, reanalysed as el arte (masculine). The feminine nouns ave and hambre, however, which exhibit the same morphology as arte when appearing with the feminine el, complicate the seemingly transparent historical development of the change in gender from el arte (f.) > el arte (m.). The present study seeks to identify factors beyond ambiguous gender marking that can explain why a change in gender from feminine to masculine has occurred in the case of el arte, but not in the case of el ave and el hambre, thus contributing, not only to our understanding of this one aspect of Spanish historical grammar, but also to issues of gender and gender change in general linguistic theory.

Summary Although Spanish philologists have long been aware of the origin of the first person plural imperative of the verb ir, ¡Vamos!, (as well as reflexive ¡Vámonos!), none has even remarked on the synchronic irregularity of the affirmative-negative pattern vamos ~ no vayamos vis-à-vis that of all other verbs, e. g., cantemos ~ no cantemos, comamos ~ no comamos, salgamos ~ no salgamos, etc., in which the same form appears in both the affirmative and negative. Nor has anyone recognized that the original pattern was indeed regular, i. e., vayamos ~ no vayamos (and vayámo(s)nos ~ no nos vayamos). It therefore remains to be explained how, when, and why this original, regular affirmative-negative structure was replaced by the irregular structure now found in Modern Spanish. The present study will attempt to answer these questions through a detailed diachronic morphosyntactic analysis.

El estudio de la morfología en la historia del español, i. e., la morfología histórica española, juega un papel principal en la gramática histórica española (i. e., la historia "interna" del español) y, por consiguiente, es un componente de suma importancia de la historia del español en general. Aunque el cambio fonético ha ocasionado tremendos cambios en la lengua a través de los siglos, muchas estructuras del español actual (y antiguo) son el resultado del cambio morfológico. Por ejemplo, mientras el cambio del lat. habuī gtentity esp. ant. ove (esp. mod. hube) se explica por los siguientes cambios fonéticos regulares: habuī → [habwi gtentity abwi gtentity au̯bi gtentity ou̯bi gtentity oβe], lat. tenuī nunca hubiese cambiado al español antiguo tove sólo por el cambio fonético. Más bien, tenuī hubiese dado una forma como **teune, así: tenuī → [tenwi gtentity teu̯ni gtentity teu̯ne]. Así que, cuando el cambio fonético regular no explica la evolución de una forma, es útil considerar la posibilidad de que algún tipo de cambio morfológico haya ocurrido.

Spanish feminine nouns beginning with tonic /a/ are preceded by the definite article el, rather than la, e.g., el agua, el hambre, etc. This el is a diachronic continuation of the Latin feminine demonstrative ILLA, whose final /a/ merged with the initial tonic /a/ of the nouns concerned, e.g., ILLA AQUA > ela agua > el agua. Thus, from a historical perspective, this el is feminine, and has therefore been known in the scholarly literature as the "feminine el". However, given that most native speakers of Spanish are unfamiliar with the history of their language, they misinterpret this el to be the masculine definite article used in this context to avoid a perceived cacophonous sequence of two identical vowels. This erroneous reinterpretation of feminine el as masculine has led many speakers to use other masculine modifiers before these feminine nouns, e.g., este agua, poco agua, todo el agua, etc. This use of other masculine modifiers before these nouns has led some scholars to reclassify them as "hermaphroditic", i.e., half masculine, half feminine, since masculine modifiers can appear before, but only feminine modifiers after. The present study introduces a syntactic criterion, previously overlooked, which intends to resolve the issue.

The informal singular irregular imperatives of Spanish have traditionally been explained as the result of primarily phonological processes. However, such an approach presents various problems which to date are unresolved: DIC > di, but FAC > faz > haz, not *fa > *ha. VENI > ven, without diphthongization because of -I, but TENE > ten, not *tien. Ten has been explained as analogical to ven, but a form like *vien could just as well have resulted from analogy with the phonologically expected *tien. Also, SEDE > se, without expected diphthongization, a form that cannot be explained as analogical to ven. Finally, pon, sal, ten, ven, (val) have been explained as resulting from apocope during the Middle Ages, but the process is not attested in the imperative (i.e., *pone > *pon(e) > pon, etc.) as it is in the present indicative (e.g., pone > pon(e) > pon). Furthermore, it has not been satisfactorily explained why apocope was permanent in these imperatives but not in the present indicative (and other tenses). In this study, all irregularities, synchronic and diachronic, are explained through a new analysis based primarily on a process of historical morphology, namely, back-formation.

Most Spanish nouns are masculine or feminine, but a few are ambiguous, e.g., el/la azucar, el/la mar, etc. Of these, ambiguous azucar is unique because fem. azucar may appear with el, e.g., el azucar morena. The RAE suggests that this feminine el is a residual case of the OSp. fem. el which appeared before feminine nouns beginning with any vowel, e.g., el arena (now restricted to the context of tonic /a/, e.g., el agua). However, azucar was originally masculine, such that el of el azucar morena cannot be a diachronic continuation of the OSp. fem. el. Here el azucar morena-type structures are explained through a historical analysis involving processes of contamination and morpho-phonological reanalysis.

The informal singular irregular imperatives of Spanish have traditionally been explained as the result of primarily phonological processes. However, such an approach presents various problems which to date are unresolved: d?c > di, but f?c > faz > haz, not *fa > *ha. v?n? > ven, without diphthongization because of ?, but t?n? > ten, not *tien. Ten has been explained as analogical to ven, but a form like *vien could just as well have resulted from analogy with the phonologically expected *tien. Also, s?d? > se, without expected diphthongization, a form that cannot be explained as analogical to ven. Finally, pon, sal, ten, ven, (val) have been explained as resulting from apocope during the Middle Ages, but the process is not attested in the imperative (i.e., *pone > *pon(e) > pon, etc.) as it is in the present indicative (e.g., pone > pon(e) > pon). Furthermore, it has not been satisfactorily explained why apocope was permanent in these imperatives but not in the present indicative (and other tenses). In this study, all irregularities, synchronic and diachronic, are explained through a new analysis based primarily on a process of historical morphology, namely, back-formation.

One of the most challenging topics of Spanish historical phonology, and perhaps of all of the history of the Spanish language, has to be the fate of the word-initial consonant represented in Latin with the letter F*whatever its phonetic reality was (i.e., whether a voiceless labiodental [f] or bilabial [8] fricative)*which, in most words, first became a voiceless glottal fricative [h], and later, in many dialects of Spanish, including the standard, disappeared altogether. This voiceless glottal fricative was first represented in Old Spanish with the letter f, then later with the letter h, which, with the loss of [h], one might say, went silent. This series of phonetic, phonological, and orthographic changes is shown with one example in (1a), (1b) and (1c):

The present study analyzes both diachronically and typologically the synchronically irregular Spanish syntagm, entre tú y yo “between you and me”, which employs the subject pronouns, tú and yo, rather than the expected prepositional pronouns, ti and mí. First, a thorough examination of Old Spanish texts reveals that the Old Spanish syntagm did indeed exhibit prepositional pronouns, i.e., OSp. entre mí & ti (a fact virtually unknown to specialists of Spanish historical grammar), thus unveiling a syntactic change from entre mí & ti > entre tú y yo. Next, the change is compared to the typologically similar ongoing change in Modern English, namely, between you and me > between you and I. It is then shown here that although the change in these languages appears, on the surface, to be an exact parallel, the origin, or cause, of the change in each language is different. Nevertheless, certain factors — including the ultimate inability of the preposition to govern the second of two conjoined elements, as well as word-order patterns of the subject pronouns — were indeed found to have played a role in the development of both entre tú y yo and between you and I.

ROMANCE LINGUISTICS: AN EVOLVING DISCIPLINE Joel Rini University ofVirginia From my perspective as a practicing Romance linguist who concentrates primarily on Spanish historical grammar, I would say that, like the languages we study, ours is an evolving, not a dying discipline . I have certainly seen changes in my short life as a Romanist. For example, when I was a graduate student at the University of Michigan not all that long ago (1982-1987), the question of the usefulness of comparative reconstruction was a highly debated topic. Some believed it was an absolutely necessary practice, since Classical Latin, representing only one register oflanguage, and a written one at that, could not serve as a starting point for our diachronic linguistic inquiries , and that one must therefore reconstruct Proto-Romance. Others believed that such a practice was absolutely futile because the majority of the reconstructed forms could be found in Classical Latin, and others perhaps in a slightly altered form. For example, no /-m/ on nouns (if, that is, they did indeed descend from the Latin accusative case), regularized first person singulars and infinitives in the case of deponent verbs, such as *morio for morior, *morire for mori, etc. Some even referred to Classical Latin as a "cheat sheet" which could be consulted to check the validity of the reconstructed form. So they asked: "Why do it at all?" Yet others believed that the comparative reconstruction of ProtoRomance should be practiced, not despite the existence of Classical Latin, but precisely because of its existence, and therefore its ability to "validate" the reconstructed forms, serving in turn as a model for the practice of comparative reconstruction of even older languages . The reconstruction of Proto-Romance as supported by the recorded Classical language somehow "proved" that other reconstructed La corónica 31.2 (Spring, 2003): 89-95 90Joel RiniLa corónica 31.2, 2003 proto-languages, such as Proto-Indo-European, and the communities in which such proto-languages would have been spoken, must have also existed. Specialists of general historical linguistics still cite Latin and Romance forms in their quest to reconstruct Proto-Indo-European , Nostratic, and the Mother Tongue.1 But the practice of comparative reconstruction of Proto-Romance has now been totally abandoned by Romance linguists, and, in effect, died out, at least in the United States, with the passing of Robert A. Hall, Frederick B. Agard, and Clifford S. Leonard. But although we no longer work backwards in time to reconstruct Proto-Romance, the formal study begun in the nineteenth century of the unfolding of Latin into modern languages, remains alive, even if only to varying degrees, depending on the Romance language concerned . But even if it is true that there are many more practitioners, at least in North America, ofSpanish historical grammar, than ofFrench or Italian, specialists of the history of Spanish will often benefit from at least a glance at some other Romance data. Paul Lloyd's magnum opus, From Latin to Spanish (1987), for example, despite the limited purpose reflected in its title, is chock full of comparative Romance data. And to cite the most recent example I could find, Thomas D. Cravens has published Comparative Historical Dialectology. Italo-Romance clues to Ibero-Romance sound change (2002), in which he adduces ItaloRomance data to attempt to resolve long-standing problems of IberoRomance . In some of my own work, comparative Romance data has been indispensable. For example, positing the chronology of changes from Latin MEtXM > Sp. conmigo would not have been possible without evidence from Old Portuguese mego and comego, Old Italian meco, Northern Italian mego, con mego, and con meo, Modern Florentine con meho, and Sardinian cum megus. The question of why the contraction oí Ieél > IeI was impeded in Spanish creer, leer,poseer andproveer, but notseer (> ser) and veer (> ver), would not have been answered without asking why Ieél > lèi was not impeded in any of the corresponding Portuguese forms. In the reduction of Latin habeo to Spanish he (as well the parallels in Western Romance, Portuguese hei, French ai, etc.), pan-Romance data confirm an intermediate *ayo, before its reduction to he. The anomalous outcome of Latin ciconia as Spanish cigüe...