Scholar, interested in language change, its history and underlying mechanisms, with a special focus on Frisian as a unique case to study these phenomena Address: Amsterdam, Noord-Holland, Netherlands
Bilingual speakers store cognates from related languages close together in their mental lexicon... more Bilingual speakers store cognates from related languages close together in their mental lexicon. In the case of minority languages, words from the dominant language often exert influence on their cognates in the minority language. In this article, we present a model describing that influence as a function of frequency and of (dis)similarity (representational distance). More specifically, it is claimed that the strength of the influence of one word upon another is (among others) a function of their frequencies divided by their formal dissimilarity. The model is applied to the distribution of nouns derived from adjectives in Frisian, where the suffix -ens competes with -heid. Of these two suffixes, Frisian -heid is similar to Dutch -heid, whereas Frisian -ens does not have a similar counterpart in Dutch. The model predicts that Frisian derived nouns of which the adjectival bases are similar in form and meaning to Dutch will occur more often with -heid and less often with -ens. It also predicts that this effect will be stronger as the words involved are more frequent. Our findings make it possible to verify the model's quantification of the influence of Dutch words on their cognates.
Keywords: bilingualism, representational convergence, language modelling, word storage, analogy, blocking
Interference Frisian (IF) is a variety of Frisian, spoken by mostly younger speakers, which is he... more Interference Frisian (IF) is a variety of Frisian, spoken by mostly younger speakers, which is heavily influenced by Dutch. IF exhibits all six logically possible word orders in a cluster of three verbs. This phenomenon has been researched by Koeneman & Postma (2006), who argue for a parameter theory which leaves frequency differences between various orders unexplained. Rejecting K&P’s parameter theory but accepting their conclusion that Dutch (and Frisian) data are input for the grammar of IF, we will argue that the word order preferences of speakers of IF are determined by frequency and similarity. More specifically, 3-verb clusters in IF are sensitive to:
(i) their linear left-to-right similarity to 2-verb clusters and 3-verb clusters in Frisian and in Dutch
(ii) the (estimated) frequency of 2- and 3-verb clusters in Frisian and Dutch.
The model will be shown to work best if Dutch and Frisian, and 2- and 3-verb clusters have equal impact factors. If different impact factors are taken, the model’s predictions do not change substantially, testifying to its robustness. This analysis is in line with recent ideas that the sequential nature of human speech is more important to syntactic processes than commonly assumed, and that less burden need be put on the hierarchical dimension of syntactic structure.
O. Strik: Modelling Analogical Change A history of Swedish and Frisian verb inflection. 165-177., 2015
There are interesting analogical patterns to be found in the history of inflection class change —... more There are interesting analogical patterns to be found in the history of inflection class change — and stability — in Frisian. It has become clear that the analogical models were better, overall, at predicting the stability of verbs than at predicting the correct direction of inflection class shift: in all cases, the proportion of correct predictions was higher for the verb systems as a whole than for the subgroup of verbs that had historically undergone shift.
This leads us to conclude that analogy by itself — as modelled by the Analogical Modeling program (AM) and the Minimal Generalization Learner (MGL) — does not possess the full conservative force needed to explain historical patterns of stability. Analogy can account for the majority of cases where verbs remained stable, but still predicted that a minority would change, when this was not the case. In other words, it is a bit too eager to reorganise the system. Another candidate for explaining diachronic stability is token frequency, particularly in that a higher token frequency is thought to make forms more resistant to morphological change, and therefore to inflection class shift. In this chapter, we present a pilot study that relates the results of analogical modelling to the token frequency of the verbs involved.
We aim to demonstrate that agent-based models can be a useful tool for historical linguists, by m... more We aim to demonstrate that agent-based models can be a useful tool for historical linguists, by modeling the historical development of verbal cluster word order in Germanic languages. Our results show that the current order in German may have developed due to increased use of subordinate clauses, while the English order is predicted to be influenced by the grammaticalization of the verb 'to have'. The methodology we use makes few assumptions, making it broadly applicable to other phenomena of language change.
Linguistic Approaches to Bilingualism 5:3, pp. 357-379, 2015
In standard Dutch, the plural suffix -en is homographic and homophonic with the linking suffix -e... more In standard Dutch, the plural suffix -en is homographic and homophonic with the linking suffix -en (boek+en “books”, boek+en+kast “bookcase”), both being pronounced as schwa. In Frisian, there is neither homography nor homophony (boek+en “books’, pronounced with syllabic nasal; boek+e+kast “bookcase”, pronounced with a linking schwa). Seeing that many areas of Frisian grammar are subject to interference from Dutch, we investigated whether Frisian-Dutch bilinguals exhibit interference from Dutch with respect to the linking suffix during Frisian speech production. Two types of Frisian-Dutch bilinguals emerged: Speakers who had Dutch as their first language tended to maintain the Dutch system of homophony between plural and linking suffixes when speaking Frisian, by using the Frisian plural as a linking morpheme. Speakers who had Frisian as their first language often maintained the Frisian system of no homophony when speaking Frisian. The implications for morphological theories are discussed.
Bilingual speakers store cognates from related languages close together in their mental lexicon... more Bilingual speakers store cognates from related languages close together in their mental lexicon. In the case of minority languages, words from the dominant language often exert influence on their cognates in the minority language. In this article, we present a model describing that influence as a function of frequency and of (dis)similarity (representational distance). More specifically, it is claimed that the strength of the influence of one word upon another is (among others) a function of their frequencies divided by their formal dissimilarity. The model is applied to the distribution of nouns derived from adjectives in Frisian, where the suffix -ens competes with -heid. Of these two suffixes, Frisian -heid is similar to Dutch -heid, whereas Frisian -ens does not have a similar counterpart in Dutch. The model predicts that Frisian derived nouns of which the adjectival bases are similar in form and meaning to Dutch will occur more often with -heid and less often with -ens. It also predicts that this effect will be stronger as the words involved are more frequent. Our findings make it possible to verify the model's quantification of the influence of Dutch words on their cognates.
Keywords: bilingualism, representational convergence, language modelling, word storage, analogy, blocking
Interference Frisian (IF) is a variety of Frisian, spoken by mostly younger speakers, which is he... more Interference Frisian (IF) is a variety of Frisian, spoken by mostly younger speakers, which is heavily influenced by Dutch. IF exhibits all six logically possible word orders in a cluster of three verbs. This phenomenon has been researched by Koeneman & Postma (2006), who argue for a parameter theory which leaves frequency differences between various orders unexplained. Rejecting K&P’s parameter theory but accepting their conclusion that Dutch (and Frisian) data are input for the grammar of IF, we will argue that the word order preferences of speakers of IF are determined by frequency and similarity. More specifically, 3-verb clusters in IF are sensitive to:
(i) their linear left-to-right similarity to 2-verb clusters and 3-verb clusters in Frisian and in Dutch
(ii) the (estimated) frequency of 2- and 3-verb clusters in Frisian and Dutch.
The model will be shown to work best if Dutch and Frisian, and 2- and 3-verb clusters have equal impact factors. If different impact factors are taken, the model’s predictions do not change substantially, testifying to its robustness. This analysis is in line with recent ideas that the sequential nature of human speech is more important to syntactic processes than commonly assumed, and that less burden need be put on the hierarchical dimension of syntactic structure.
O. Strik: Modelling Analogical Change A history of Swedish and Frisian verb inflection. 165-177., 2015
There are interesting analogical patterns to be found in the history of inflection class change —... more There are interesting analogical patterns to be found in the history of inflection class change — and stability — in Frisian. It has become clear that the analogical models were better, overall, at predicting the stability of verbs than at predicting the correct direction of inflection class shift: in all cases, the proportion of correct predictions was higher for the verb systems as a whole than for the subgroup of verbs that had historically undergone shift.
This leads us to conclude that analogy by itself — as modelled by the Analogical Modeling program (AM) and the Minimal Generalization Learner (MGL) — does not possess the full conservative force needed to explain historical patterns of stability. Analogy can account for the majority of cases where verbs remained stable, but still predicted that a minority would change, when this was not the case. In other words, it is a bit too eager to reorganise the system. Another candidate for explaining diachronic stability is token frequency, particularly in that a higher token frequency is thought to make forms more resistant to morphological change, and therefore to inflection class shift. In this chapter, we present a pilot study that relates the results of analogical modelling to the token frequency of the verbs involved.
We aim to demonstrate that agent-based models can be a useful tool for historical linguists, by m... more We aim to demonstrate that agent-based models can be a useful tool for historical linguists, by modeling the historical development of verbal cluster word order in Germanic languages. Our results show that the current order in German may have developed due to increased use of subordinate clauses, while the English order is predicted to be influenced by the grammaticalization of the verb 'to have'. The methodology we use makes few assumptions, making it broadly applicable to other phenomena of language change.
Linguistic Approaches to Bilingualism 5:3, pp. 357-379, 2015
In standard Dutch, the plural suffix -en is homographic and homophonic with the linking suffix -e... more In standard Dutch, the plural suffix -en is homographic and homophonic with the linking suffix -en (boek+en “books”, boek+en+kast “bookcase”), both being pronounced as schwa. In Frisian, there is neither homography nor homophony (boek+en “books’, pronounced with syllabic nasal; boek+e+kast “bookcase”, pronounced with a linking schwa). Seeing that many areas of Frisian grammar are subject to interference from Dutch, we investigated whether Frisian-Dutch bilinguals exhibit interference from Dutch with respect to the linking suffix during Frisian speech production. Two types of Frisian-Dutch bilinguals emerged: Speakers who had Dutch as their first language tended to maintain the Dutch system of homophony between plural and linking suffixes when speaking Frisian, by using the Frisian plural as a linking morpheme. Speakers who had Frisian as their first language often maintained the Frisian system of no homophony when speaking Frisian. The implications for morphological theories are discussed.
NOWELE. North-Western European Language Evolution, 2018
Dying language varieties are often characterised by attrition of their structural features and la... more Dying language varieties are often characterised by attrition of their structural features and large scale borrowing from the competing dominant language. The linguistic characteristics of such a dying language are particularly difficult to interpret when an extinct language is only scarcely attested (Trümmersprache) and its major informant is a non-native, who learned the language from native speakers who were potentially only imperfect learners. The East Frisian Harlingerland dialect is attested only in a manuscript booklet from the late 17th century (1691), written by the local priest Johannes Cadovius-Müller, who was not a native speaker of Frisian. He uses the two infinitive markers of Frisian in an unhistorical way. This study tries to disentangle whether this is the result of his individual imperfect learning of the language or of a restructuring in the language of the natives speakers. It is hypothesised that the last generation of native speakers of the dialect developed a synchronic phonological rule for the distribution of the two infinitive markers, based on vowel harmony, which was again partly distorted by Cadovius-Müllers imperfect learning.
The text known as 1643a from the corpus of early-Modern West Frisian texts is here reinterpreted ... more The text known as 1643a from the corpus of early-Modern West Frisian texts is here reinterpreted as a specimen of the Frisian language that existed until the 17 th century in North Holland. While it is widely acknowledged that Frisian was the language of Holland from the early Middle Ages, where it was gradually replaced by Dutch from the 9 th century, the language itself was only known through indirect evidence of place names, substratum words and a few glosses. This text is the first coherent example of Frisian spoken in Holland. Although no author and/or location is mentioned, the Hollandic origin can be identified on the basis of a number of features: its place of publication (Amsterdam) and the fact that it was addressed to young women in Amsterdam; the content of the verse, describing a practice of courting common in North Holland in the 17 th century; its linguistic features which are on the one hand clearly Frisian, yet differ substantially from West Frisian as attested in the 17 th century; and finally because of many linguistics features associated with the Dutch dialects of North Holland in the 17 th-20 th century. 1. Ynlieding Yn it korpus ier-Nijfrysk stiet in tekst mei it nûmer 1643a (Feitsma 1956: 58-59), dy't ek opnommen is yn de Taaldatabank Frysk (Taaldatabank 1.0). It giet om in fers mei as titel "MYn Liaf is eensoo swieten dier", dat ano-nym ôfprinte is yn "d' Amsteldamsche/ MINNE-ZUCHJENS", in bondel lieten, printe yn Amsterdam. De tekst is opfallend om't er in soad wurden befettet dy't op dy wize yn gjin inkelde Westerlauwersk Fryske tekst foar-komme, wylst guon foarmen en wurden oerienkomsten met it Noardhol-lânsk sjen litte. Dat is de oanlieding om te ûndersykjen oft wy hjir net mei in fers te krijen hawwe kinne dat skreaun is yn it Frysk fan Noard-Hollân. Of-Us Wurk, jiergong 67 (2018), s. 1-37
Frisians and their North Sea Neighbours, from the fifth century to the viking age., 2017
The West Germanic language group is traditionally divided into North Sea Germanic (Old English an... more The West Germanic language group is traditionally divided into North Sea Germanic (Old English and Old Frisian) and Continental Germanic (Old Low Franconian and Old High German). The position of Old Saxon within this West Germanic language continuum is ambiguous: on the one hand, it shares various innovations with Old English and Old Frisian, on the other, it has many traits in common with Old High German. The North Sea Germanic characteristics are more abundant in the minor Old Saxon texts, especially those associated with Eastphalia than in Heliand, Genesis and other sources from the region of Essen and Werden. This implies that the ‘coastal’ features were more widespread in the Old Saxon dialects of the deep south-east inland than in the dialects of the south-western regions, closer to the coast. This rather unexpected distribution of linguistic characteristics is the focus of the present paper, which offers a detailed mapping of a range of North Sea Germanic features in the Old Saxon material. The patterns resulting from the mapping confirm the assumed dialectal contrasts in Old Saxon. The aim of the paper is to account for the attested distribution of linguistic features. The explanation is sought in the ancient traffic networks which, given the poor overland traffic conditions, overlapped with river basin networks. More specifically, the central and south eastern regions stayed in direct contact with the Frisian-speaking region on the North Sea littoral through the rivers Ems, Weser and Elbe, while the south-western regions were in close contact with the continentally-oriented Franconians in the Rhine Valley. The inland extension of the North Sea Germanic linguistic features can therefore be interpreted as a reflection of communication networks, which developed along the lines of economic and political contacts, and was directed by the travel networks of those days.
Amsterdamer Beiträge zur älteren Germanistik 77: Festschrift Rolf Bremmer, 2017
Rolf Bremmer (2007) concludes that the language of the Old Frisian Riustring manuscripts shows tr... more Rolf Bremmer (2007) concludes that the language of the Old Frisian Riustring manuscripts shows traces of copying from texts written in other Old Frisian dialects, notably from the Ems region. The strongest indication for his hypothesis comes from the masculine plural ending-ar, which is the rule in Ems Old Frisian but the exception in R1 and absent from other Riustring manuscripts. In this contribution, Bremmer's hypothesis is partly confirmed, but augmented with the reconstruction of an indigenous Riustring plural ending-ar in masculine a-stem nouns denoting an animate subject, which appear substantially more often in the nominative. Nouns with a higher frequency of occurrence in the accusative take the plural ending-a. This is taken to reflect a former Proto-Frisian situation, with the ending-ar in the nom. pl. of masculine a-stem nouns against-a in the acc. pl., similar to Old Norse. The earlier distribution had become lexicalised by the time of Riustring Old Frisian. Some of the attested instances, however, are better explained as remnants of a copying process from Ems Old Frisian.
Proto-Germanic (PGmc.) ai in stressed syllables shows varied outcomes in Germanic languages (ā, ē... more Proto-Germanic (PGmc.) ai in stressed syllables shows varied outcomes in Germanic languages (ā, ē, ei), with many of these developments being conditioned by different phonological contexts. This article presents a reconstruction that unifies this variation by assuming that the monophthongisation spread over 'Germania' in two waves with different scopes and directions. The first wave expanded from north to south and was primarily limited to the contexts before-h and-r. A second wave, affecting the remaining instances of PGmc. ai, did not reach Old High German and Old West Nordic. The whole process covered the time between 400 and 900. The monophthongisation of PGmc. ai does not reflect any structural contrast among the Germanic languages, but the results had a differentiating impact on their vowel systems. The presented reconstruction is consistent with the information from runic inscriptions. It supposes a geographical configuration of tribes in a post-Migration setting.
The Old Frisian verbal ending-ath (pl.pres. and 3 rd sg.) appears either as <ath> or <at>. Van He... more The Old Frisian verbal ending-ath (pl.pres. and 3 rd sg.) appears either as <ath> or <at>. Van Helten's assumption that they reflect pre-'Verner' stress alternation seems too far-fetched. The analysis of the verbal endings in the oldest Old Frisian sources, R 1 , B 1 , E 1 , H 2 and the Old Skeltariucht in Unia, reveals various phonological factors, with a clear east-west divide, grouping R 1 and E 1 together against U and H 2. The form-ath is the archetype (91% in B 1). In the east,-at is a reflex of the emergence of the un-marked [t] instead of the marked [θ] in contexts with a low acoustic salience, as it appears preferably at the end of a phrase before a pause. In the western manuscripts, the variant-ath is best preserved following light roots. Long-distance dissimilation in e.g. werthath > werthat is found in most of the texts. These phonological conditions are superimposed upon a chronological gradual west-east directed transition-ath >-at. A statistical model covering the data from the five manuscripts renders their (approx-imate) geographical location, age and the respective phonological conditions (different for east and west!) statistically significant, with geography being the most important factor. Together, they provide a 70% accurate prediction of the alternation between-ath and-at on the level of indivudual tokens in the four texts.
Amsterdamer Beiträge zur älteren Germanistik, 2016
The recent discovery of a few snippets with Old Frisian words from the 12th century was a sensati... more The recent discovery of a few snippets with Old Frisian words from the 12th century was a sensational surprise for the study of Old Frisian (Langbroek 2015). The 8 words are linguistically interpreted according to the inflectional categories of the parallel Latin glosses. This is in line with the wide spread custom in mediaeval interlineair glossing practices. This article argues that Langbroek's linguistic interpretation, which turns lesa 'to redeem' into a subjunctive or imperative and herte 'heart' into a plural form, is linguistically not sustainable. It is postulated that all word forms, except for mi, reflect lemma base forms, i.e. infinitives or nominative singular forms. This may pose new questions about the character of these glosses and their application in the original context and use. Keywords Old Frisian-psalm glosses 1 Einführung Die Überlieferung des Altfriesischen wurde letztes Jahr erfreulicher weise um einige Psalmglossen bereichert, die auch noch sehr alt sein sollen, nämlich aus dem 12. Jahrhundert (Langbroek 2015). Das ist um ein Jahrhundert älter als die sonstige Überlieferung des Altfriesischen. Es handelt sich zwar nur um 8 Wörter, wovon zweimal das gleiche Wort, mi ‚mir/mich', dennoch ist für eine Sprache wie das Altfriesische jegliche Information wertvoll.
The unaccented syllables of Runic Frisian (6th-9th c.) have been studied on several occasions (e.... more The unaccented syllables of Runic Frisian (6th-9th c.) have been studied on several occasions (e.g. by Nielsen and Nedoma), but such studies have failed to offer a comprehensive analysis and reconstruction of the phonological system as a whole. Building upon the insights gained in earlier studies, all unstressed vowels are investigated here in a coherent treatment. The resulting hypothesis about the Runic Frisian unstressed vowel system is that it was very similar to the systems of nearly contemporaneous Old Saxon and early Old English. Previous analyses agree that the use of –u in the nom.sg. of masc. a-stems (ka[m]bu ‘comb’) is a Runic Frisian peculiarity. It is considered to reflect a /ə/ < Proto-Germanic *-a-. The use of word-final –a (instead of * e or * æ) in deda, Old Frisian dede ‘did’ is one of the most problematic instances among the less-ambiguous inscriptions. Seemingly inexplicable spelling alternation in etymologically identical vowels can be understood if we assume that Runic Frisian /æ/ (Old Frisian <e>) was rendered by <a, æ> while Runic Frisian /å/ (Old Frisian <a>) was rendered by <a, o>. The same spelling practice has been identified in Old Saxon.
Amsterdamer Beiträge zur älteren Germanistik, 2016
The Old Saxon ending of the NAp of the masculine a-stems is generally-os in the larger sources He... more The Old Saxon ending of the NAp of the masculine a-stems is generally-os in the larger sources Heliand and Genesis, appears as-as in minor texts, but faces competition from-a in minor texts in southwestern sources. There are various hypotheses about the origin of the-a-ending, whether going back to the original Proto-Germanic accusative ending or being a borrowing from Franconian. An analysis of all 119 attested NAp forms of masculine-(j/w)a-stems from minor texts reveals that the ending-a arose in the late 9th century and found its roots indeed in the southwestern region of the Lower Rhine Valley (Essen, Werden). The variant-os, next to geographically dominant-as, is only found in that region as well. The ending-a most likely spread from Lower Rhine Valley Franconian, where an intense contact existed between Franconian and Saxon, the former being the politically dominant variety. This borrowing was the first step in a total reshaping of Middle Low Saxon plural morphology, based on the Franconian pattern of masculine-e versus feminine-en. Keywords Old Saxon-nominal declension-language contact-dialects 1 Einleitung Die meisten männlichen Hauptwörter in den germanischen Sprachen gehören der Klasse der a-(und ja-und wa)-Stämme an, wie altsächsisches (as.) dag, Nominativ und Akkusativ Plural (NAp) dagos. Die historische Bewertung der
H. Bloemhoff & N. Streekstra: Basisboek historische taalkunde van het Nederlands: met afzonderlijke aandacht voor het Brabants, Fries, Limburgs, Nedersaksisch, Vlaams en Zeeuws, 329-340, 2015
Those legal terms in the Lex Frisionum of the late 8th century that are introduced by the phrase ... more Those legal terms in the Lex Frisionum of the late 8th century that are introduced by the phrase ‘quod uocant’ or ‘quod dicunt’ are generally considered to be Frisian. In this study, the phonological and morphological character of these words is analysed and compared to evidence from the contemporaneous Frisian runic corpus and later Old Frisian. It turns out that none of the fourteen words show an unambiguously Frisian shape, while various instances lack features that are expected to have been present in the Frisian language of that time. It is therefore concluded that the fourteen words cannot be considered as Frisian from a phonological and morphological perspective. They are most likely Low Franconian. This means a severe reduction of the linguistic evidence for Frisian from that early period. In a lexicological sense, the words can very well be Frisian and probably originate in the Frisian legal terminology of those days.
All East- and North Frisian dialects and all the Old East Frisian sources attest to the Old Frisi... more All East- and North Frisian dialects and all the Old East Frisian sources attest to the Old Frisian ending –ar in the nom./acc.pl. of the masculine a-stems. For Old West Frisian, current grammars of Old Frisian mention only the existence of the ending –an in this paradigm form, which may have replaced the older –ar ending before the first attestations. The archaic Old West Frisian texts in Codex Unia reveal seven instances of –ar as a nom./acc.pl. ending of masculine nouns in. These instances attest to the earlier existence of –ar also in Old West Frisian.
An additional indication that in pre-attested Old West Frisian the plural ending –ar was categorically replaced by an, comes from the historical s-stem words klāth ‘cloth’ and kind ‘child’, which also show up with –an in Old West Frisian as well, against–ar in Old East Frisian. In an excursus to Middle and early-Modern West Frisian it is shown that the contrast between uninflected plurals for historical a-stem neuter nouns and –an in the s-stem nouns left its traces until ca. 1700.
Our knowledge of Old Germanic languages, just as of other ‘old’ languages, relies entirely on the... more Our knowledge of Old Germanic languages, just as of other ‘old’ languages, relies entirely on the materials that have come down to us over the centuries. For various socio-historical and cultural reasons, the amount of the material for the earliest attested stages of Germanic is limited, and the paucity of available data delimits the amount of detail that can be deduced, inevitably distorting our knowledge of the linguistic facts.
By using the token-lemma curves we were able to conclude that the variation in text type has a substantial impact on the number of attested lemmas in Old English and Old Frisian. The lack of specific lemmas in Old Frisian can largely be ascribed to the remarkably homogenous nature of its corpus, comprising primarily legal texts. However, the differences between Old English and Old Frisian seen at the lexical level cannot be interpreted as an a priori difference between the two language realities and their lexicons. Both corpora, despite clear discrepancies in their sizes and composition, turn out to be large enough to draw reliable conclusions about genreindependent features, such as phonology or morphology: by comparing the Old Frisian corpus to both the entire Old English corpus and the Old English legal texts sub-corpus, it was demonstrated that this type of information is basically genre-independent and is not affected by the corpus size. Such a conclusion seems to be fairly encouraging and promising for any comparative linguistic studies involving an investigation into older, scarcely attested language stages.
Amsterdamer Beiträge zur älteren Germanistik 72, pp. 35-62., 2014
This article offers a phonological analysis of the vowel system of Runic Frisian (6th-9th c.) in ... more This article offers a phonological analysis of the vowel system of Runic Frisian (6th-9th c.) in stressed syllables. By using the most reliable attestations and taking them at face value, a vowel system emerges that can be the ancestor of all the later attested Frisian varieties, including the rather deviating Insular North Frisian dialects. The position of PGerm. *ē1 remains unclear due to scarcity of the data. In the short vowel system, the most outstanding features are the separation of fronted PGerm. *a from PGerm. *e (as is also the case in Old English). The most outstanding outcome in the long vowel system is the development of PGerm. *ai. Current econstructions treat its monophthongisation as one process, contemporaneous with or even preceding the monophthongisation of PGerm. *au. This analysis proposes an early date of monophthongisation of PGerm. *ai in velar contexts and of PGerm. *au, and a later date of monophthongisation of PGerm. *ai in other contexts. The emigration of Frisians to the North Frisian islands in the 7th/8th century is positioned between those two stages. This reconstruction solves a long-standing problem in the historical phonology of Frisian, most prominent in the developments in Insular North Frisian and reconciles the main-stream opinions with a chronology of events as proposed by Hofmann (1964).
Descriptions of late-19 th century Germanic dialects suggest or even explicitly claim that they w... more Descriptions of late-19 th century Germanic dialects suggest or even explicitly claim that they were shaped in the late Middle Ages or the Early Modern period. This implies that 19 th century dialects represented a language that had been nearly frozen for at least 300 years. The presumed stable character of the dialects was deliberately confirmed in the 20 th century by selecting old and conservative informants. This archaic perception of the 19 th century dialects is difficult to verify due to the lack of reliable dialectal data from the preceding centuries. West Frisian seems to be a sole exception to this. A moderate but fairly continuous flow of Frisian texts since ca. 1400, written in a language that did not develop a standardized form until the 19 th century, opens a window on 600 years of language history. A series of 10 case studies from various linguistic domains shows an ongoing dynamism over the centuries and no sign of early 'frozen' dialects. The changes are either the result of language contact with Dutch or language internal pressure and the geographical sources and directions of change reflect the shifting configurations in economy and demography through the centuries.
Twenty-nine smiles for Alastair: Freundesgabe für Dr. Alastair G. H. Walker zu seinem Abschied von der Nordfriesischen Wörterbuchstelle der Christian-Albrechts-Universität zu Kiel am 4. Juli 2013, 2013
De herbewoning van de Friese kwelders en terpnamen Een onderzoek naar mogelijke verbanden 1. Inle... more De herbewoning van de Friese kwelders en terpnamen Een onderzoek naar mogelijke verbanden 1. Inleiding en probleemstelling De kwelderzône langs de zuidelijke Noordzee laat een sterke afname van de bevolking zien in de 3 e eeuw n.Chr., gevolgd door een grootschalige hervestiging van bewoners vanaf het einde van de 4 e eeuw/begin van de 5 e eeuw (Boeles 1951, Nicolay 2005, Nieuwhof 2016). 1 De gebieden die volgens archeologische of historische bronnen door een Germaans(talig)e bevolking bewoond geraken gedurende de periode die traditioneel als de 'Volksverhuizingstijd' aangeduid wordt, zijn rijk aan plaatsnamen die gevormd zijn met het element *haim of *ing (bijv. Nielsen 2000: 307-314; Berkel & Samplonius 2018: s.v. heem, ing). Het eerste element betekent zoveel als 'huis, hof, dorp', het tweede duidt op een '(toebe)horen aan/bij'. Met dergelijke elementen gevormde namen bevatten als eerste element vaak een persoonsnaam. De namen die met *haim gevormd zijn, verschijnen in de Friese gebieden 2 als eindigend op-um: Warkum, Jellum, Hantum, Mullum; de namen met *ing hebben een typische Friese verandering doorgemaakt en eindigen tegenwoordig opens: Battens, Hesens, Tettens. Naast deze twee naamtypen is er een derde veelvoorkomend naamvormend element in de Friese kuststreken: *wurþ, dat in de huidige namen verschijnt in de vorm van-werd/t,-weerd,-ward,-aard: Holwert, Feerwerd, Fallward, Hidaard. Oorspronkelijk betekent ook dit element waarschijnlijk meer algemeen 'boerderij, huisplaats' (Berkel & Samplonius 2018: s.v. werd), maar het is in de context van de Lage Landen steeds nadrukkelijk verbonden aan woonheuvels of terpen en nadrukkelijker dan de *haim-en *ing-namen verbonden met het historisch-geografische landschap. De vraag die hier onderzocht wordt, luidt: is er een verband tussen de naamtypen enerzijds en de geografische en chronologische spreiding van de nieuwe nederzettingen in het Friese kweldergebied? De 0-hypothese is dat de *haim-namen mogelijk net iets ouder zijn en dus oververtegenwoordigd zouden zijn onder de namen van de oudste nederzettingen, terwijl *ing wellicht een tweede fase zou kunnen markeren, net als de *wurþ namen, uitgaande van-ondertussen grotendeels achterhaalde-ideeën over regressie-en transgressieperioden: herbevolking gedurende een regressieperiode, met vervolgens terpenbouw (en dus *wurþ-namen) gedurende een volgende transgressieperiode. Zo zijn er opvallende opeenhopingen van-ens en-wert/-aard-namen in zuidelijk Westergo met zwaardere, nattere kleibodems. Ook blijkt er een oververtegenwoordiging van-ens-namen in het Jadegebied te zijn (Versloot 2011: 132-134), iets dat betekenisvol zou kunnen zijn gegeven het feit dat Ostfriesland in zijn algemeenheid wat later bewoond geraakt is dan Fryslân en Groningen (Behre 2014: 123). Geografische variatie van naamtypen kan dus globaal zijn: van west naar oost, maar ook regionaal: van hoge kwelder richting nattere delen (knipkleigebieden) en verder het veen in, elk mogelijk gekoppeld aan de tijd van bewoning. 2. Materiaal en analyse Voor een zinvolle analyse is een grote gegevensverzameling nodig. Die is aanwezig in de vorm van Tabel 6 in het proefschrift van Egge Knol (1993: 94-103). 3 Weliswaar is dit materiaal al wat ouder, 1 Was dit gegeven bij de eerste ontdekking ervan door Boeles in het begin van de 20 e eeuw nog iets bijzonders, intussen is vastgesteld dat in veel gebieden van het vroeger 'Germania' sprake is van verregaande ontvolking in bijv. de 3 e eeuw ten zuiden van de Limes (Roymans, Heeren & Derks 2020), de 5 e en 6 e eeuw in delen van Noord-Duitsland (Ludowici 2019), enz. 2 Hiermee is het Middeleeuwse Friesland bedoeld: in deze studie ligt de nadruk op het gebied tussen Vlie en Ley. 3 Ik dank Egge Knol en Annet Nieuwhof hartelijk voor hun hulp bij het vinden en interpreteren van het archeologische materiaal en hun commentaar op een eerdere versie.
Word order changes and grammaticalization in Germanic verbal clusters
In this work, we model t... more Word order changes and grammaticalization in Germanic verbal clusters
In this work, we model the historical development of verbal cluster order in Germanic languages. While there is an ongoing debate on the syntactic structure of these clusters, we created a simple model of surface patterns in which we view each order as a separate outcome, with a probability distribution over the outcomes. This type of modeling lets us explore the diverging development of verbal clusters in these languages, taking a reconstruction of the state of verbal clusters in Proto-Germanic as a starting point. The models converge from their manually defined, Proto-Germanic initial probability distribution to a state in which probabilities are distributed based on the features of the model. We then compare the resulting model output with actual Germanic language texts to see how well we have modeled the real state of these languages. We show that the interaction of basic probabilistic choices of constructions with shifting input and shifting preference of constructions may be a key to understanding different word orders as observed in the Germanic languages.
This article is devoted to borrowing in Frisian of Dutch irregular plurals formed by suffixation ... more This article is devoted to borrowing in Frisian of Dutch irregular plurals formed by suffixation and stem vowel lengthening; e.g. dak [dak] ‘roof’ - daken [da:k?/da:kn?] ‘roofs’ instead of Frisian dak [dak] - dakken [dakn?]. Irregular plural formation is unproductive in Dutch and uncommon in Frisian (except in a couple of historical borrowings). Although the appearance of irregular Dutch plurals in Frisian had not previously been described or investigated, we hypothesize their existence in spoken Frisian, based on analogical cases. However, they would not be borrowed without any limitations as various factors would affect their borrowability; e.g. words frequency, phonotactic constraints or paradigm alternation. This study is based on approximately 175 interview recordings. The participants were asked to name different objects shown to them during the interview picture task. The results confirm both the (quite frequent) use of Dutch irregular plurals in Frisian and the involvement o...
A number of runic inscriptions from the entire Germanic area from between A.D. 200 and 800 exhibi... more A number of runic inscriptions from the entire Germanic area from between A.D. 200 and 800 exhibit non-etymological, epenthetic vowels, such as worahto for *worhto ‘did’. An analysis of all (likely ...
Bei der aus den skandinavischen Sprachen und den weserfriesischen Mundarten bekannten ‘Vokalbalan... more Bei der aus den skandinavischen Sprachen und den weserfriesischen Mundarten bekannten ‘Vokalbalance’ handelt es sich um eine unterschiedliche Entwicklung der unbetonten Vokale auf Grund der Quantitat der vorausgehenden Stammsilbe. Im vorliegenden Aufsatz wird an erster Stelle gezeigt, dass die Unterschiede in der Entwicklung der unbetonten Silben eher als quantitativ denn als qualitativ zu bewerten sind, womit die Vokalbalance als Ganzes als eine quantitative Begebenheit gekennzeichnet werden kann. Anhand von einer zweifachen Entwicklung im Festlandnordfriesischen des kurzen afr. /e/ in offener Silbe, wobei sich afr. /e/ in offener Silbe vor folgendem /a/ der Entwicklung in geschlossener Silbe anschliest, lasst sich darauf eine regressive Form von Vokalbalance nachweisen. Allerdings hatte Lofstedt (1928, 1933) bereits Ahnliches beschrieben, aber dies wurde von diesem Autor anders bewertet. Der auch wiederum quantitativ zu bewertende Unterschied zwischen /a/ und /e/ im Altfriesischen...
The Old Frisian verbal ending -ath (pl.pres. and 3rd sg.) appears either as or . Van Helten’s ass... more The Old Frisian verbal ending -ath (pl.pres. and 3rd sg.) appears either as or . Van Helten’s assumption that they reflect pre-‘Verner’ stress alternation seems too far-fetched. The analysis of the verbal endings in the oldest Old Frisian sources, R1, B1, E1, H2 and the Old Skeltariucht in Unia, reveals various phonological factors, with a clear east-west divide, grouping R1 and E1 together against U and H2. The form -ath is the arche-type (91% in B1). In the east, -at is a reflex of the emergence of the un-marked [t] instead of the marked [θ] in contexts with a low acoustic salience, as it appears preferably at the end of a phrase before a pause. In the western manuscripts, the variant -ath is best preserved following light roots. Long-distance dissimilation in e.g. werthath > werthat is found in most of the texts. These phonological conditions are superimposed upon a chronological gradual west-east directed transition -ath > -at. A statistical model covering the data from th...
Disclaimer/Complaints regulations If you believe that digital publication of certain material inf... more Disclaimer/Complaints regulations If you believe that digital publication of certain material infringes any of your rights or (privacy) interests, please let the Library know, stating your reasons. In case of a legitimate complaint, the Library will make the material inaccessible and/or remove it from the website. Please Ask the Library: http://uba.uva.nl/en/contact, or a letter to: Library of the University of Amsterdam, Secretariat, Singel 425, 1012 WP Amsterdam, The Netherlands. You will be contacted as soon as possible.
... Wol wurde parallellen yn oare wurden neamd. De betsjutting is rûnorn dy fan guod (benanunen p... more ... Wol wurde parallellen yn oare wurden neamd. De betsjutting is rûnorn dy fan guod (benanunen planten, reid, kroas ensfh.) dat op it wetter driuwt en/of op it strän oanspielt. ... 5 Page 6. bleaun wie, Iykas yn it Aldfrysk.&quot; In Hollànsk toponym as Torp (by Den Helder) ~ Wlfr. terp &lt; Afr. ...
Uploads
Keywords: bilingualism, representational convergence, language modelling, word storage, analogy, blocking
(i) their linear left-to-right similarity to 2-verb clusters and 3-verb clusters in Frisian and in Dutch
(ii) the (estimated) frequency of 2- and 3-verb clusters in Frisian and Dutch.
The model will be shown to work best if Dutch and Frisian, and 2- and 3-verb clusters have equal impact factors. If different impact factors are taken, the model’s predictions do not change substantially, testifying to its robustness. This analysis is in line with recent ideas that the sequential nature of human speech is more important to syntactic processes than commonly assumed, and that less burden need be put on the hierarchical dimension of syntactic structure.
This leads us to conclude that analogy by itself — as modelled by the Analogical Modeling program (AM) and the Minimal Generalization Learner (MGL) — does not possess the full conservative force needed to explain historical patterns of stability. Analogy can account for the majority of cases where verbs remained stable, but still predicted that a minority would change, when this was not the case. In other words, it is a bit too eager to reorganise the system. Another candidate for explaining diachronic stability is token frequency, particularly in that a higher token frequency is thought to make forms more resistant to morphological change, and therefore to inflection class shift. In this chapter, we present a pilot study that relates the results of analogical modelling to the token frequency of the verbs involved.
Keywords: bilingualism, representational convergence, language modelling, word storage, analogy, blocking
(i) their linear left-to-right similarity to 2-verb clusters and 3-verb clusters in Frisian and in Dutch
(ii) the (estimated) frequency of 2- and 3-verb clusters in Frisian and Dutch.
The model will be shown to work best if Dutch and Frisian, and 2- and 3-verb clusters have equal impact factors. If different impact factors are taken, the model’s predictions do not change substantially, testifying to its robustness. This analysis is in line with recent ideas that the sequential nature of human speech is more important to syntactic processes than commonly assumed, and that less burden need be put on the hierarchical dimension of syntactic structure.
This leads us to conclude that analogy by itself — as modelled by the Analogical Modeling program (AM) and the Minimal Generalization Learner (MGL) — does not possess the full conservative force needed to explain historical patterns of stability. Analogy can account for the majority of cases where verbs remained stable, but still predicted that a minority would change, when this was not the case. In other words, it is a bit too eager to reorganise the system. Another candidate for explaining diachronic stability is token frequency, particularly in that a higher token frequency is thought to make forms more resistant to morphological change, and therefore to inflection class shift. In this chapter, we present a pilot study that relates the results of analogical modelling to the token frequency of the verbs involved.
The East Frisian Harlingerland dialect is attested only in a manuscript booklet from the late 17th century (1691), written by the local priest Johannes Cadovius-Müller, who was not a native speaker of Frisian. He uses the two infinitive markers of Frisian in an unhistorical way. This study tries to disentangle whether this is the result of his individual imperfect learning of the language or of a restructuring in the language of the natives speakers. It is hypothesised that the last generation of native speakers of the dialect developed a synchronic phonological rule for the distribution of the two infinitive markers, based on vowel harmony, which was again partly distorted by Cadovius-Müllers imperfect learning.
An additional indication that in pre-attested Old West Frisian the plural ending –ar was categorically replaced by an, comes from the historical s-stem words klāth ‘cloth’ and kind ‘child’, which also show up with –an in Old West Frisian as well, against–ar in Old East Frisian. In an excursus to Middle and early-Modern West Frisian it is shown that the contrast between uninflected plurals for historical a-stem neuter nouns and –an in the s-stem nouns left its traces until ca. 1700.
By using the token-lemma curves we were able to conclude that the variation in text type has a substantial impact on the number of attested lemmas in Old English and Old Frisian. The lack of specific lemmas in Old Frisian can largely be ascribed to the remarkably homogenous nature of its corpus, comprising primarily legal texts. However, the differences between Old English and Old Frisian seen at the lexical level cannot be interpreted as an a priori difference between the two language realities and their lexicons. Both corpora, despite clear discrepancies in their sizes and composition, turn out to be large enough to draw reliable conclusions about genreindependent features, such as phonology or morphology: by comparing the Old Frisian corpus to both the entire Old English corpus and the Old English legal texts sub-corpus, it was demonstrated that this type of information is basically genre-independent and is not affected by the corpus size. Such a conclusion seems to be fairly encouraging and promising for any comparative linguistic studies involving an investigation into older, scarcely attested language stages.
In this work, we model the historical development of verbal cluster order in Germanic languages. While there is an ongoing debate on the syntactic structure of these clusters, we created a simple model of surface patterns in which we view each order as a separate outcome, with a probability distribution over the outcomes. This type of modeling lets us explore the diverging development of verbal clusters in these languages, taking a reconstruction of the state of verbal clusters in Proto-Germanic as a starting point. The models converge from their manually defined, Proto-Germanic initial probability distribution to a state in which probabilities are distributed based on the features of the model. We then compare the resulting model output with actual Germanic language texts to see how well we have modeled the real state of these languages. We show that the interaction of basic probabilistic choices of constructions with shifting input and shifting preference of constructions may be a key to understanding different word orders as observed in the Germanic languages.