¹ Consult especially Tomás Navarro, Aurelio Espinosa (hijo), and L. Rodríguez-Cas-tellano, “La frontera del andaluz,” RFE, xx (1933), 225–378; Pedro Henriquez Urena El problema del andalucismo dialectal de America (Buenos Aires, 1932); Amado Alonso, “Problemas de dialectologîa hispanoamericana,” in Vol. i of the Biblioteca de Dialectologia Eispanoamericana (BDH), (Buenos Aires, 1930); and his Estudios lingulsticos: lemas his-panoamericanos (Madrid, 1953). The latter volume includes reprints of the following important studies: “La base lingüística del espanol americano,” “Origenes del seseo americano,” “La ll y sus alteraciones en España y América,” and “-r y -l en España y America.”

² See Pedro Benvenutto Murrieta, El lenguaje peruano (Lima, 1936), pp. 118–119; also Juan de la Riva-Aguero's introduction to Roberto Levillier's Audiencia de Lima (154964), (Madrid, 1922), tomo i. “Andalucia y Extremadura plasmaron el Perú” (p. xvii), etc.

³ “… si Andalucía dió alguna vez predominio de conquistadores y colonizadores, eso tuvo que ser en los primeros tiempos, y justamente la América de los primeros treinta años se redujo al Caribe, y más concretamente, a las islas” (Alonso, Temas hispanoamericanos, p. 327).

⁴ See Navarro's introduction to his El español en Puerto Rico (Univ. de Puerto Rico, 1948). Says Padin, Rev. Est. Eisp., I, 51, “Yerra también Mixer (en su libro Porto Rico, Nueva York, 1926) al hablar del origen andaluz del gibaro (el campesino puertorriqueño). Casi todos los primeros pobladores que fueron a Puerto Rico procedían de las dos Castillas.”

⁵ For a discussion of the climatic theory and the argument it provoked between the scholars Max Leopold Wagner and Pedro Henríquez Urena, consult Wagner, “El supuesto andalucismo de América y la teoría climatológica,” RFE, xiv (1927), 20–32, and Henríquez Urena, Sobre el problema del andalucismo dialectal de America (Buenos Aires, 1932), pp. 121 ff., 129 ff.

⁶ The growth of the Spanish-American dialects has varied in each region with a number of factors, each of them important but none in itself decisive. These include: (1) the dates of conquest and the regional composition of the conquistadores; (2) the population density and cultural level of the conquered Indians; (3) the varying degrees of bilingualism and linguistic interchange arising from such results of the conquest as religious conversion, mestizaje, enslavement, extinction, the importation of Negroes from Africa; and (4) (something all too often forgotten) the contributions of subsequent waves of colonists and the amount of cultural contact throughout the colonial period with Spain itself.

⁷ Luis Rubio y Moreno, Pasajeros a Indias (Sevilla, 1917), published in Col. docs, inéds. Para la hist, de Iberoamérica, Vols, viii and xiii; Pedro Henriquez Urena, pp. 1–118 (entitled Comienzos del espanol en America); Juan Rodriguez Arzúa, “Las regiones espanolas y la población de America (1509–38),” Rev. Indias (Madrid), xxx (1947), 695–748; V. Aubrey Neasham, “Spanish Emigrants to the New World: 1492–1592,” HAHR (1950), pp. 147–160. See also the articles by Juan Friede, “The Catdlogo de pasajeros and Spanish Emigration to 1550,” HAHR (1951), pp. 333–348, and “Algunas observaciones sobre la realidad de la emigración española a America en la primera mitad del siglo xvi” (Rev. Indias, Madrid, xlix [1952], 467–496). Miss Vivian M. Gruber, apparently unfamiliar with the study by Arzúa, has in the Florida State Univ. Stud., No. m (1951), 1–7, a short note, “Peninsular Origins of Spain's First American Colonists,” which sums up the tabulated figures of Henríquez Urefia, Neasham, Icaza, and (for Chile) Thayer Ojeda.

⁸ Angel Rosenblat's masterly study, La poblaci ón ind ígena de America desde 1492 hasta la aclualidad (Buenos Aires, 1945), calculates that in 1570 there were 140,000 white persons living in the Spanish American colonies. How many of these were native-born criollos it is hard to estimate, but allowing for these on the one hand, and on the other for mortalities and continued immigration until 1600, I would suggest that slightly over 200,000 is not an unreasonable estimate of the total number of Spanish settlers up to the end of the 16th century.

⁹ For the Antillean period my chief sources of information were, besides the Catálogo de pasajeros a Indias (Sevilla, 1940, Vol. i), the 67-volume Colección de documentas inédilos de Indias (Pacheco, Cárdenas, Torres de Mendoza et al.) for which Ernesto Schäfer has compiled a valuable indice alfab êtico de personas (îndice de la colecciôn de documentos inéditos de Indias, Madrid, 1946, tomo i); the Colección de documentos inéditos para la historia de Iberoamérica, Vols, v, x, xi, xiv; José Toribio de Medina, El descubrimiento del Océano Pacijko (Vols, i and ii, Santiago de Chile, 1913–14), which deals with Balboa and Pedrarias; Francisco A. de Icaza, Diccionario autobiográfico de conquistadores y pobladores de Nueva España, 2 vols. (Madrid, 1923); Alonso Dorantes de Carranza, Sumaria relación de las cosas de la Nueva España (1604), Mexico, 1902; Agustín Millares Carlo, índice y extractos de los Protocolos del Archivo de Notarias de Mexico, D.F., tomo I, 1524—28 (El Colegio de Mexico, 1945); and the historians and chroniclers Bernai Diaz del Castillo, Historia verdadera de la conquista de la Nueva España, Gonzalo Fernandez de Oviedo, Historia general y natural de las Indias, Fray Bartolomé de las Casas, Historia de las Indias, Juan de Castallanos, Elegías de varones ilustres de Indias, and Francisco López de Gómara, Historia de la Conquista de México. For identifying place names in Spain, I used principally Pascual Madoz's monumental 16-volume Diccionario geográfico-estadístico-histórico de España y sus posesiones de uitramar (Madrid 1845–50).

¹⁰ See Alonso, Tentas hispanoamericanos, “La base lingüística del espanol americano.”

¹¹ I have chosen the year 1520 because with it starts a new phase in the Spanish colonization of the New World: the assault upon the mainland. With the opening up for settlement of Mexico, Venezuela, Central America, New Galicia, Peru, the Plate region and the New Kingdom of Granada, a flow of emigration began to those places that originated not so much in the Antilles as in Spain itself. However, the initial landings in Mexico by Cortes in 1519 and Narváez in 1520 may still be considered as terminating the preceding (Antil-lean) period, inasmuch as the members of these two expeditions were recruited entirely from among the island settlers.

¹² The Kingdom of Granada, comprising the present-day Andalusian provinces of Almería, Granada, and Malaga, had at the dawn of the colonial period only just been conquered and was in many respects not yet part of Andalusia proper. The statistics for the Antillean period show this division very well:

¹³ ángel Rosenblat, p. 81, estimates that even as late as 1570 Puerto Rico's 3 towns boasted a total of only 200 vecinosor 1,000 whites.

¹⁴ Not until a captain's fame had reached Spain did he attract to his standard large numbers of his paisanos. Later we may find an Hernando Pizarro and an Hernando de Soto depleting the towns of Extremadura with their irresistible call to seek fame and fortune in the Indies. (Out of 60S who followed de Soto on his ill-fated expedition to Florida in 1538 no less than 269 [44%] were from his native Badajoz, with 50 of these men from his native town alone!) But in the early period I must emphasize that this is not the case. Diego Velazquez Ponce de León, Cortés, Ojeda, Balboa, Pedrarias, all commanded expeditions in which their compatriots were in the minority.

¹⁵ Cartas de Eugenio de Salazar, vecino y natural de Madrid, escritas a muy particulares amigos suyos, publicadas por la Sociedad de Bibliófilos Españoles (Madrid, 1866), pp. 3537. See on this subject A. Alonso, Temas cit., pp. 63–67; also Berta Elena Vidal de Battini, Voces marinas en el habla rural de San Luis, in Filologia, (Buenos Aires), I, 105–149.

¹⁶ Historia general …, ed. Amador de los Rios (Madrid, 1851–55), n, 473 ff.

¹⁷ M. M. de Peralta, Costa Rica, Nicaragua y Panama en el siglo xvi, pp. 527–539, gives a “Sumaria descriptión del Reyno de Tierra Firme, Uamado Castilla del Oro, que está subjeto a la Real Audiencia de Panamá, por el Dr. Alonso criado de Castilla, Oidor decano de la misma. Nombre de Dios, 7 de mayo de 1575,” wh'ch reads in part: “la ciudad de Panamá … tendrá quatrocientas casas … en que habrá quinientos vezinos, y de hordinario asisten ochocientos nombres poco más o menos. Es la gente muy política, todos españoles y gran parte dellos originarios de la ciudad de Sevilla. Es gente de mucho en-tendimiento; su oficio es tratar y contratar, ecepto quince o veynte vezinos que tratan los campos y viven de los ganados y hacienda que en ellos tienen. Es por la mayor parte gente rica.… En este pueblo está la gente con poco asiento y como de camino para passar al Perú o venir a Espana. Es mucho el comercio y trato desta ciudad, ansí de las cosas de Espana como del Perú por estar en medio de las dos mares del Norte y del Sur, y muy acomodada para las contrataciones.”

¹⁸ For a description of Seville in the 16th century, see the historical work of Santiago Montoto, Sevilla en el Imperio (Siglo, xvi) (Sevilla, 1938). Interesting documentary evidence may also be found in the Archiva de Protocoles de Sevilla (Vols, v, x, xi, and xiv of the CDI) and in the 3 volumes of the Catáiogo de Pasajeros a Indias.

¹⁹ Because some of the pasajeros a Indias would register as vecinos de Sevilla after residing in Seville for only 1 or 2 years, and since the location of the registry in Seville made registration easier for residents of Seville than for other people, I have made every legitimate effort to offset such statistical advantages. In the first place, the very incomplete Catáiogo de pasajeros has been supplemented with numerous other sources, mostly colonial, in which Seville could not conceivably enjoy any statistical advantage. Next, though I have not attempted to differentiate them statistically, my files observe the following categories of identification: natural, vecino, hijo de vecinos, and hijo de naturelles, identifications which sometimes by implication involve the family accompanying one so described. In general, the category natural takes precedence over all others. Seville has been stripped of several vecinos for whom I could find birthplaces elsewhere. Cases of dual vecindad (e.g., “vecino de Sevilla o de Lebrija”) would on principle be decided in favor of the smaller town, the latter being the more likely origin in view of the general drift towards the big cities.

When in 1529 a royal decree authorized other Spanish ports to traffic with the Indies, it is probable that many northern emigrants then found it convenient to sail directly from Coruna or Laredo without registering in Sevilla, but in the Antillean period this had not yet become a disturbing factor.

Careful study of the surnames of registered pasajeros of known origin has revealed for the years 1510–19 a positive correlation of from 70% to 85% (average: 75.9%) between the town, province, or region of actual origin and 772 distinguishing appellations of the type Juan Rodriguez de Cuéllar, Alonso de Badajoz or Pedro Vizcaíno. This correlation is much higher for persons of humble or rustic origin than it is for persons of exalted rank or profession, as the former traveled little and some had no surname at all until they departed for America. The correlation is also higher for persons bearing the names of places located in the South, because during the course of reconquista and especially with the discovery of America the direction of migration of future colonists was almost invariably from north to south and not the other way. Discreet application of this correlation has permitted identification, with a high degree of probability, of certain very early cases where no other evidence may ever be available. But of the small percentage of control cases in which the correlation breaks down, nearly half involve legitimate sevillanos with surnames suggestive, however, of places outside Andalusia. This means then that in the few cases where persons have been identified only by their names, Seville has suffered a slight statistical penalty. But despite all these penalties and safeguards, Seville still emerges as incomparably the heaviest single contributor to the initial phase of Spanish colonial effort. Weighing all factors together I believe that my ratio of semllanos to other colonists is, for the Antillean period, substantially correct.

²⁰ For documentations of the superior prestige in the 16th and early 17th centuries of the courtly speech of Toledo, see esp. Amado Alonso's Castettano, español, idioma nacional. Historia espiritual de très nombres (Buenos Aires: Institute de Filología, 1938, and Buenos Aires: Losada, 1943 and 1949. See pp. 67–72 and 91–95 of the Losada editions). The speech of Andalusia, insofar as it differed from the then emerging national idiom, enjoyed considerably less prestige north of its own border than it does today.

²¹ Though it is true that a person's speech is originally formed by the environment in which he is raised, it can be altered partially or even completely by a new environment with which he may seek to associate himself for reasons of real or fancied prestige. Americans returning from England with acquired Oxford accents exemplify this very well.

²² Enterprising Andalusian pilots, captains, and seamen early transferred their base of operations to Santo Domingo and Havana in order to engage in the inter-island traffic and man the ships needed by the island colonists for their expeditions of discovery, conquest, and trade.

²³ When 4 years ago the late Amado Alonso encouraged me to undertake this study I had no preconceived ideas about what I would discover, though the earlier statistical studies of Pedro Henriquez Urena, Arzúa, and others did seem to indicate that by the end of the 16th century the accumulated contributions of the “North” and “South” of Spain were about equal. My task was to reorganize and supplement their work in such a way that it could be used for regional and chronological linguistic studies and would reveal definite migration patterns between certain regions in certain years or decades.

Such a pattern has clearly revealed itself for the initial period, thus far lending illusory historical evidence to the popular notions about the Andalusian origin of “American Spanish,” at least as far as the Antilles are concerned. But we must constantly bear in mind that the numerical preponderance of Andalusians was only one of several factors that helped shape the first Antillean dialect and only one of a great many more that helped form the Antillean dialects of today. These statistics are merely guides, pointing, in cases of substantial majorities or of group migrations in certain years, to regional speech trends which may either have been reinforced or neutralized, then or later, by other developments. Taken by themselves, these statistics have no linguistic validity. But properly used as supporting evidence by historians of language and society, they can be of no little help in answering the all-important questions of who, where, when, and how.

²⁴ When a colonist's home town was known but not the year of emigration, I used for statistical purposes the earliest year for which there was positive evidence of his presence in America; e.g., on several of the original expeditionaries to Mexico I had no information prior to 1519, ‘20, or even ‘21. However in order to have sailed with Cortéz they must have been in Cuba in 1518, so this became the earliest positive date I could employ.