Hogmanay!

Vandaag is de laatste dag van het jaar 2018. Een dag die goed is voor twee posts op mijn blog, want verder is er weinig anders te doen dan eten, drinken en samenzijn.

In Schotland, een van de landen die ik dit jaar heb bezocht en waar ik wederom (net als bij Japan) met een goed gevoel een aantal dagen heb vertoefd, heet de laatste dag van het jaar Hogmanay. Deze laatste dag gaat gepaard met gewoontes als het geven van cadeautjes en het bezoeken van buren en vrienden.

Het woord hogmanay zou volgens linguïsten een Gaelic, Franse en Noorse oorsprong kunnen hebben. Ik denk dat het anders zit, namelijk als volgt:

Het woord hogmanay is afgeleid van een Oudnederlands* woord, namelijk higunnan.Dit woord, dat bekend is uit het regeltje

Hebban olla uogala nestas hagunnan hinase hi(c) (a)nda thu uuat unbidan uue nu

betekent zoiets als beginnen.

Dat zou heel goed kunnen aansluiten bij het idee dat na oudjaarsavond het volgende jaar begint.

Hogmanay! Een nieuw begin. Bouwt allen een nieuw nest in 2019!

© Rick Ruhland 2019

* Terzijde, volgens sommige literatuurvorsers annex taalwetenschappers is deze regel niet geschreven in een oudere versie van ons huidige Nederlands, maar heeft de tekst een Oudwestnederfrankische oorsprong, en misschien zelfs een Kentse (als in het graafschap in Engeland), waar naar alle waarschijnlijkheid de schrijver van de regel,  een West-Vlaamse monnik, Latijnse teksten overschreef in een klooster.

Advertisements

Scenes from a thesis: competence & performance

“The idea of an underlying structure is a crucial element of generative grammars. This idea makes it plausible that language is more than just a string of sounds and sentences. Furthermore, complementary evidence for underlying structures can be found in a discussion on competence and performance. When adults are asked to, they can perfectly well say which utterances are and which utterances are not grammatical. Apparently, we humans have inbuilt knowledge of language, usually referred to as competence. Even children (at the age of 4) seem to have a knowledge of language when they are offered incorrect sentences that they have to repeat (Ruhland, 1991)*. In daily life, we do not use all language knowledge we can [use], for practical reasons. Performance is what we use everyday in communication: it is our competence hampered by factors like memory and abilities of the hearer or the reader to understand language.”

© Rick Ruhland 2018

* De acquisitie van ‘hoeven’ / Rick Ruhland’, In: Tabu. Jaargang 21, 1991.

Scenes from a thesis: final remarks

My Ph.D. thesis Going the distance was published 20 years ago.

The only reason to dedicate a couple of blogs on that thesis was that science should not stay within the walls of research institutes and universities.

Okay, there are two reasons. The second reason for paying some attention to the thesis is that I think my findings are still of great value and truth. It is good research.

Those are the two reasons why…

No, there are three reasons. It is 20 years ago – anybody sense a Beatles reference here? Could be: the thesis is full of references to pop and rock music – this year since I got my Ph.D. Which I never use anymore (I am not a scientist anymore).

So, there you are. These are the main three reasons (for any Monty Python lovers reading this post, there is a reference to a famous sketch).

If you like to read more, here is a link to the thesis. If you have questions about the research, the thesis, the bigger picture, do not hesitate to contact me.

© Rick Ruhland 2018

WoW: woordspelling

Woordgrappen maken zonder te weten hoe je iets schrijft; soms ook: taalgrap die een kreun bij het publiek oproept omdat de grap eigenlijk te voor de hand liggend is. Linguïstisch: weten welke letters er in een woord horen, maar niet snappen dat er ook een grappige connotatie aan verbonden is.

 

© Rick Ruhland 2018

Scenes from a thesis: Relationship between growth and structural models

“[First,] The structural model and the growth model may be completely autonomous. The growth model is capable of describing the observed changes in performance without recourse to hypotheses on the quality of the developmental process that may follow from a structural model. In fact, this seems to imply that the growers do not entertain any relation to the notions of the structural model. Thus, the growth model neither supports, nor refutes the assumptions of the structural model.

Second, growth models and a linguistic theory are compatible. This comes in two varieties. First, it may be the case that the parameters of a growth model that was constructed inductively (i.e. with the single objective of obtaining a perfect fit with the data) can be meaningfully interpreted in terms of a structural model, and that this leads to (quantitative) predictions that concur with the predictions derived from the structural model. Alternatively, a growth model that was constructed deductively, i.e. by choosing its parameters on the basis of notions and considerations supplied by a structural model, is supported by the empirical data.

Third, growth and linguistic models may contradict each other. Again, we envision two versions of this situation. In the top-down scenario, a growth model that was informed by a structural model does not fit the data, whereas, possibly, a bottom-up (i.e. based on the data) model does. Alternatively, an inductively constructed model comprises parameters that can be interpreted as refutations of (some of) the core assumptions of the structural model. In either case, the bottom-up model (that was validated by its fit on the data only) may be thought of as the instantiation of a hitherto unidentified structural model.”

From: Going the distance:  A Non-Linear Approach To Change In Language Development / H.G. Ruhland. Groningen, 1998.

© Rick Ruhland 2018

Proefschrift van de week

Titel: Versprekingen: de motor van taalevolutie. Auteur: Jinke Flongen.

Samenvatting: De linguïstiek is niet wat je noemt een vakgebied dat zich vaak in de kijker speelt. Vaak zijn toepassingen van taalwetenschappers maar beperkt geldig. Althans, die indruk wekken linguïsten mijns inziens meestal. Misschien hebben zij met hun bevindingen wel meer in hun mars, en is de impact hun wetenschappelijke vragen en inzichten en uitkomsten veel groter dan taalwetenschappers zelf beseffen.
Dat gezegd hebbend: niet elk proefschrift op het gebied van de linguïstiek is dus de moeite waard. Ik lees af en toe proefschriften die van los zand en speeksel aan elkaar hangen. Flingen is daarvan een flagrant voorbeeld. Hij komt met een bulk aan voorbeelden (waarvan niet duidelijk is of hij die zelf heeft verzameld of dat die uit een bestaand corpus komen). De voorbeelden die hij aanhaalt zijn zowel fonologisch, morfologisch, syntactisch, als semantisch van aard. Het is een grote vergaarbak. Ratjetoe. Radjetoe.

Eindoordeel: Ik denk dat Flongen uit alle macht een reden zocht naar DE reden hoe en waarom taal zich ontwikkelt. Anders dan een linguïst – die elke verhandeling mathematisch dan wel logisch aanpakt en veranderingen en fouten (en wat Ties meer zei) als een gevolg van de taal zelf ziet – is Flongen de weg ingeslagen van probabilistische taalwetenschap. Wat ‘zou kunnen’ is belangrijker als verklaringsmodel dan een theorie. Ik denk dat versprekingen best mee spelen in taalverandering, maar dit lijkt meer een kwestie van ‘de kok horen luiden, maar niet weten waar de lepel hangt’.

© Rick Ruhland 2018

Scenes from a thesis: explanations for the change

“Possible explanations fo the sudden change in function words is sudden.

First, the function words increase due to an increase in other variables (i.e. the change is caused by other independent factors). The best candidate presumably is the amount of words spoken during a session. Analyses, however, suggest that the words per session do increase over time (the production of speech increases as the child becomes older), but there is no coherence between this independent variable and the sudden change in function words. Thus, the increase of words per session is probably not the cause of the increase of function words. It might be the other way around: the productivity increases due to an increase in function words and other variables. The problem is a ‘chicken and egg’ problem. It is likely that much of the increase of the productivity is caused by the introduction of function words.

Second, the increase of utterance length would call for functional categories. That is, the lengthening of the utterances (expressed by an increase in both MLU and multiword utterances) is the result of an increase of lexical categories. Since lexical categories like nouns and verbs call for functional categories, longer sentences make syntactic slots become available. Longer sentences and the availability of syntactic slots can explain the timing of appearance of function words. However, they cannot explain the suddenness of growth of function words.

Third, the parents might increase their use of function words, and children follow the parental increase. This would indicate that a child uses the input to imitate it directly or with a time lag. However, on the grounds of the analyses, this assumption must be rejected. The multiple regression analyses do not confirm this hypothesis. The increase of function words in the language of a child is solely the result of an age effect. Thus, timing nor shape of change are the result of a change in language of the parents.”

From: Going the distance:  A Non-Linear Approach To Change In Language Development / H.G. Ruhland. Groningen, 1998.

© Rick Ruhland 2018

Scenes from a thesis: fits and averages

“In sum, there is no evidence that the development of function words is best explained from a linear or discontinuous point of view. […] The non-linear fitting has implications for the study of language development. […] It does not mean that age averages are wrong.
However, one must be careful with respect to certain claims about development. Group data do not yield a solid description of (individual) development. Only age dependent development (i.e. strictly timed development) will benefit from such averages. The second point to be made is that whatever the averages, in all cases the data are best modelled by a non-linear fit.”

From: Going the distance:  A Non-Linear Approach To Change In Language Development / H.G. Ruhland. Groningen, 1998.

© Rick Ruhland 2018

Scenes from a thesis: syntactic categories

“Languages have a structure which is commonly referred to as grammar. This implies that language is a rule governed system. It also means that words in sentences have a relationship. Furthermore, in most languages words have a fixed order in a sentence. The assumption of an underlying structure makes it plausible that all natural languages share common features. The surface structure of languages is derived from the underlying basic order through movement procedures. The lexical categories (i.e. words) in a sentence undergo changes (e.g. verbs are inflected). The changes that form to say a shell around the lexical entities are called functional categories. Functional categories are syntactic operations, that do not refer to entities in the real world. It is assumed that these categories are more difficult to acquire.”

From: Going the distance:  A Non-Linear Approach To Change In Language Development. H.G. Ruhland. Groningen, 1998.

© Rick Ruhland 2018

Scenes from a thesis: dynamic systems models

“Dynamic systems models are not chosen on arbitrary grounds, but because they are
convenient in four ways. First, empirical data can be fitted to decide what sort of change
one is dealing with. That is, a fit with a growth model could help explain the shape of
development. Fits of developmental curves can be a first help in deciding wether or not
change is non-linear. Second, these fits have to be explained. That is, why is non-linear
behaviour present in development and especially, what are the mechanisms responsible
for this sort of change? Third, with the aid of non-linear equations and theoretical
considerations a model can be made to describe development. But this description is
more than a description if the model is correct, because, fourth, predictions can be made
on the basis of the model. So, these models have descriptive, predictive and explanatory
adequacy.”

From: Going the distance:  A Non-Linear Approach To Change In Language Development. H.G. Ruhland. Groningen, 1998.

© Rick Ruhland 2018