Joseph F. (Joseph Francis)
Eska s. xx–xxi
Works edited
Contributions to journals
As a verb-second language, one expects Middle Cornish to allow only a single argument/complement to appear in the left periphery of affirmative root clauses. Object personal pronouns never occur in the left periphery, but a full non-adjunct XP and subject personal pronoun do, in fact, coöccur in 329 clauses in our corpus—in that order, in all but a single token—, presumably owing to poetic overdetermination, which alters the morphosyntax and surface configuration in order to enable the required syllable-count or end-rhyme in the verse line. George 1990 & 1991, based upon an analysis of Beunans Meriasek, finds five tokens of full object DP and subject personal pronoun which coöccur in the left periphery, which, he states, are not motivated by poetic overdetermination. He concludes, on that basis, that the construction is generated by the grammar. In this paper, we collect all of the tokens of this construction in the verse corpus of Middle Cornish and propose that they are all, ultimately, motivated by poetic overdetermination, not only in order to enable the required syllable-count or end-rhyme, but sometimes also to encode pragmatic information.
As a verb-second language, one expects Middle Cornish to allow only a single argument/complement to appear in the left periphery of affirmative root clauses. Object personal pronouns never occur in the left periphery, but a full non-adjunct XP and subject personal pronoun do, in fact, coöccur in 329 clauses in our corpus—in that order, in all but a single token—, presumably owing to poetic overdetermination, which alters the morphosyntax and surface configuration in order to enable the required syllable-count or end-rhyme in the verse line. George 1990 & 1991, based upon an analysis of Beunans Meriasek, finds five tokens of full object DP and subject personal pronoun which coöccur in the left periphery, which, he states, are not motivated by poetic overdetermination. He concludes, on that basis, that the construction is generated by the grammar. In this paper, we collect all of the tokens of this construction in the verse corpus of Middle Cornish and propose that they are all, ultimately, motivated by poetic overdetermination, not only in order to enable the required syllable-count or end-rhyme, but sometimes also to encode pragmatic information.
This paper examines the proto‐Celtic plosive system through the lens of Laryngeal Realism. Drawing upon phonetic data from contemporary Celtic languages and philological data from medieval Insular Celtic and ancient Continental Celtic languages, it concludes that the active Laryngeal feature in these languages is not [voice], but [spread glottis], and that this feature should be projected back to proto‐Celtic. Such an analysis allows for a much more straightforward analysis of the evolution of the early Celtic plosive system, and, in particular, allows for a non‐stipulative analysis of perhaps the best known of Celtic sound changes, the loss of proto‐IE */p/, in simple aerodynamic terms. It is demonstrated, furthermore, that the loss of proto‐IE */p/ cannot be explained by contact with pre‐Basque or Iberian, but, instead, was, in all likelihood, a natural development.
This paper examines the proto‐Celtic plosive system through the lens of Laryngeal Realism. Drawing upon phonetic data from contemporary Celtic languages and philological data from medieval Insular Celtic and ancient Continental Celtic languages, it concludes that the active Laryngeal feature in these languages is not [voice], but [spread glottis], and that this feature should be projected back to proto‐Celtic. Such an analysis allows for a much more straightforward analysis of the evolution of the early Celtic plosive system, and, in particular, allows for a non‐stipulative analysis of perhaps the best known of Celtic sound changes, the loss of proto‐IE */p/, in simple aerodynamic terms. It is demonstrated, furthermore, that the loss of proto‐IE */p/ cannot be explained by contact with pre‐Basque or Iberian, but, instead, was, in all likelihood, a natural development.
[EN] M. J. Estaran Tolosa proposes in Études celtiques, XLI-2015, p. 95-109, that the form traditionally read as KuiTos in the inscription of S. Bernardino di Briona (Novara) is, instead, to be read as KuiToi and forms a syntagm with the preceding form TanoTaliKnoi. This paper argues that the final character of this form is unlike any other token of < i > in the inscription, and, indeed, it is precisely the inverse image of a token of < s > in l. A1 of the inscription. Linguistic analysis, likewise, is decidedly in favour of the traditional reading.
[EN] M. J. Estaran Tolosa proposes in Études celtiques, XLI-2015, p. 95-109, that the form traditionally read as KuiTos in the inscription of S. Bernardino di Briona (Novara) is, instead, to be read as KuiToi and forms a syntagm with the preceding form TanoTaliKnoi. This paper argues that the final character of this form is unlike any other token of < i > in the inscription, and, indeed, it is precisely the inverse image of a token of < s > in l. A1 of the inscription. Linguistic analysis, likewise, is decidedly in favour of the traditional reading.
L'auteur analyse le rapport traditionnel entre gaul. sioxti et viri, siächt et le trouve déficient aux niveaux phonologique et sémantique. A la place il propose une nouvelle interprétation basée sur la racine *seg- «attacher, toucher». Il en profite aussi pour corriger une erreur dans son ancienne analyse syntactique de l'inscription où se trouvait la forme sioxti.
[EN] The author examines the traditional equation of Gaul, sioxti OIr. si cht and finds it wanting, both on phonological and semantic grounds. In its place, he proposes a new interpretation founded on the base *seg- attach, touch’. He also takes the opportunity to correct an error in his previous syntactic analysis of the inscription in which the form sioxti occurs.
L'auteur analyse le rapport traditionnel entre gaul. sioxti et viri, siächt et le trouve déficient aux niveaux phonologique et sémantique. A la place il propose une nouvelle interprétation basée sur la racine *seg- «attacher, toucher». Il en profite aussi pour corriger une erreur dans son ancienne analyse syntactique de l'inscription où se trouvait la forme sioxti.
[EN] The author examines the traditional equation of Gaul, sioxti OIr. si cht and finds it wanting, both on phonological and semantic grounds. In its place, he proposes a new interpretation founded on the base *seg- attach, touch’. He also takes the opportunity to correct an error in his previous syntactic analysis of the inscription in which the form sioxti occurs.
I. Dernière phrase de l’inscription celtibère de Botorrita : le verbe ŕusimus, 1re p. du pl., serait un «nous de majesté», le sujet étant un individu unique dénommé par plusieurs mots, dont ComPalCo-reś, «roi du Sénat» ; II. Le verbe TośoKoTe dans l’inscription gauloise cisalpine de Verceil comporte un pron. infixe *so- plutôt que *sto-.
[EN] Two notes on Continental Celtic
I. In the last sentence of the Hispano-Celtic inscription from Botorrita, the verb ŕusimus (1st pl.) is probably a “plurale maiestatis”, the subject being a single person named with different words (particularly ComPalCo-reś “king of the Senate”). II. In the Gaulish Cisalpine inscription from Vercelli the verb TośoKoTe includes an infixed pronoun *so- rather than *sto-.
I. Dernière phrase de l’inscription celtibère de Botorrita : le verbe ŕusimus, 1re p. du pl., serait un «nous de majesté», le sujet étant un individu unique dénommé par plusieurs mots, dont ComPalCo-reś, «roi du Sénat» ; II. Le verbe TośoKoTe dans l’inscription gauloise cisalpine de Verceil comporte un pron. infixe *so- plutôt que *sto-.
[EN] Two notes on Continental Celtic
I. In the last sentence of the Hispano-Celtic inscription from Botorrita, the verb ŕusimus (1st pl.) is probably a “plurale maiestatis”, the subject being a single person named with different words (particularly ComPalCo-reś “king of the Senate”). II. In the Gaulish Cisalpine inscription from Vercelli the verb TośoKoTe includes an infixed pronoun *so- rather than *sto-.
