Non-Conditioned, Unconscious Intra-Individual Variation

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The paper addresses non-conditioned, unconscious intra-speaker variation, based on the data from the Basque linguistic atlas (EHHA). The author(s) discuss the methodology of data collection in EHHA, followed by the description and discussion of the data relevant for present research. The paper looks into several different frameworks which study variation in order to examine intra-individual variation in the EHHA data.

Data presented in this paper can be seen as an important resource for future studies in language variation. As mentioned in the paper, intra-individual variation is often neglected both in some theoretical frameworks, as well as in more empirically based branches of linguistics, like dialectology. The EHHA data was collected in such a way that as much intra-speaker variation as possible is present. The data was collected in a setting close to 'natural' speech, which makes it possible to search for variation that is not socially motivated. The methodology of data collection is well described and provides sufficient information. However, Basque area, from which the data comes from, is characterised by 'extreme' diglossic situation, with the older speakers 'suffering' from a diaglossic situation, as mentioned in the paper. This makes this dataset very specific and conclusions can be not generalised to speech in areas not characterised by such extreme diglossia. This could have been discussed in more detail in the paper since it is certainly an important factor in language use in that area.

The Results and Discussion sections are somewhat less easy to follow. In these sections the author(s) try to show that in presented cases the variation cannot be explained by certain frameworks. While the main argument is that recorded variation is always in the same sociolinguistic situation (same survey,  same field worker, in the same conversation and in a very short interval of time) and is not conditioned, for some cases, like for lexical variation, it is unclear if selected examples show intra-speaker variation or not. The author(s) try to find a plausible explanation and relate detected variation to several theoretical frameworks, but proving that something is not a specific type of variation remains difficult. Despite this, the paper can be considered a valuable contribution to the research on intra-speaker variation.

Author Response

Data presented in this paper can be seen as an important resource for future studies in language variation. As mentioned in the paper, intra-individual variation is often neglected both in some theoretical frameworks, as well as in more empirically based branches of linguistics, like dialectology. The EHHA data was collected in such a way that as much intra-speaker variation as possible is present. The data was collected in a setting close to 'natural' speech, which makes it possible to search for variation that is not socially motivated. The methodology of data collection is well described and provides sufficient information. However, Basque area, from which the data comes from, is characterised by 'extreme' diglossic situation, with the older speakers 'suffering' from a diaglossic situation, as mentioned in the paper. This makes this dataset very specific and conclusions can be not generalised to speech in areas not characterised by such extreme diglossia. This could have been discussed in more detail in the paper since it is certainly an important factor in language use in that area.

 

COMMENT: 

It is true that the diaclosic and diglossic situation undergone by the EHHA informants can be considered as important or extreme, as it can be in Galician, Catalan, in which the institutions dedicated to the recovery and unification of languages (the Academy of the Basque language or the Royal Galician Academy or the Institut d'Estudis Catalans) have made great efforts to standardise and the respective autonomous governments have in force more or less ambitious plans for the recovery and prestige of the local language; or Asturleonese or Aragonese, on another level, or Corsican and Breton, to give other examples. The informants of the EHHA meet the requirements to be considered a ‘dialect base’ in Bellmann's terminology (1998), that is, the spectrum furthest away from the standard and prestigious form of Basque. In any case, and independently of the linguistic distance there may be in Basque between the ‘dialectal base’ and the standard variety, which may be very different from one variety to the other, the diaglossic situation of Basque cannot be considered as something special, since such situations occur in all languages, whether there is greater or lesser geo- or sociolinguistic variation among speakers.

 

 

The Results and Discussion sections are somewhat less easy to follow. In these sections the author(s) try to show that in presented cases the variation cannot be explained by certain frameworks. While the main argument is that recorded variation is always in the same sociolinguistic situation (same survey, same field worker, in the same conversation and in a very short interval of time) and is not conditioned, for some cases, like for lexical variation, it is unclear if selected examples show intraspeaker variation or not. The author(s) try to find a plausible explanation and relate detected variation to several theoretical frameworks, but proving that something is not a specific type of variation remains difficult. Despite this, the paper can be considered a valuable contribution to the research on intraspeaker variation.

 

COMMENT: 

I added the following text: Regarding ‘it is unclear if selected examples show intraspeaker variation or not.’ we consider such examples to represent both intra-speaker and inter-speaker variation. If the informant does not realise that he/she pronounces the same expression in different ways or pronounces different expressions to say exactly the same thing, we consider this to be intra-speaker variation, as long as such forms have not been suggested by the fieldworker, as they have been in these cases. On the other hand, the types of variation described so far do not account for this unconscious and unconditioned variation.

 

Language changes by itself, because it is in a permanent state of flux, in which different pronunciations of the same word or expression can coexist. But not only at the phonetic or pronunciation level, but also at the morphological, syntactic or lexical level; Although some forms may be monolexematic (expressed in a single way throughout the entire geographical extent of the language, for example, in Basque the singular absolutive suffix "-a", which is common to all varieties), others may have local or/and diatopic variants. This does not mean that all the elements of the language are in a changing state at all times, but that there is always some element that produces variation and is in the process of changing. Such an element begins by producing variation and in a later state, favoured by linguistic or extra-linguistic factors, this variation may produce a change in the system; but it may also be that this variation is subsequently reduced or disappears without producing any change. However, we consider that not all elements of a language are in a state of flux at all times, but there are always a number of elements which produce variation. Moreover, it can be said that very few elements are stable at the level of pronunciation in the spoken language. On the other hand, it can be said that the greater the influence of the written language, the less variation there is in the spoken variety. The standard written language provides stability to the system and directly influences the reduction of variability. The laxity in the variability of the spoken language system allows the speaker to accommodate his or her production to personal extra-linguistic factors, which may be derived from intra- or extra-speaker factors. Casual, informal conversation is the ideal situation for producing unconditioned, unconscious variation. Contextual relaxation helps.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

This study focuses on intra-individual variation in the data from the Linguistic Atlas of the Popular Varieties of Basque on different linguistic levels. Although the data seems interesting and relevant for discussing individual linguistic repertoires, I have serious concerns about the strong claims the authors are making based on their analyses. Please also consider my extensive comments and mark-ups in the document.

In summary, my main concerns are as follows:

1) The level of awareness ('consciousness') of speakers' own variation cannot be gathered from the data as presented here. We simply do not know whether speakers realise they are using different variants or not unless you have explicit information on that. In fact, as soon as you ask informants for 'other words for the same thing', there must be some level of awareness of variation.

2) In the same vein, more detailed information on the way intra-individual variation was elicited and produced by the speakers would be needed. For example, it makes a huge difference whether a certain lexical variant was first used spontaneously by the speaker and then he/she adds "but my grandfather said XY". Clearly, then, variation is in reception rather than production and, importantly, metalinguistic comments of the latter kind indicate awareness of variation and typically give some information on constraints of variation.

3) In order to be able to argue that the kind of variation encountered is "non-conditioned" as suggested by the title and numerous times in the text, you would have to explicitly and convincingly rule out any extralinguistic AND linguistic factors potentially conditioning variation. It is true that the basic situation and the communicative partners do not change during the interview but the immediate pragmatics and linguistic contexts can and do change nonetheless! There are many instances where some linguistic conditioning may be at work - apart from social/situational constraints which may not be directly applicable but perhaps mentioned in metalinguistic observations. 

4) In my view, there is a lack of in-depth data analysis. In section 3, you give a basic quantitative overview and then a number of examples. In my view, a more extensive and thorough analysis of results is necessary. This could be quantitative or qualitative in kind, but should set out to answer specific research questions (apart from demonstrating that there IS intra-individual variation).

5) Also, more background information on the Basque language situation AND the specific variants in the examples would be necessary (e.g., what is their position in the diaglossic spectrum)?

Comments for author File: Comments.pdf

Comments on the Quality of English Language

Overall, the manuscript was comprehensible but I had trouble understanding certain passages or terms. These passages/words are marked in the document.

Author Response

This study focuses on intra-individual variation in the data from the Linguistic Atlas of the Popular Varieties of Basque on different linguistic levels. Although the data seems interesting and relevant for discussing individual linguistic repertoires, I have serious concerns about the strong claims the authors are making based on their analyses. Please also consider my extensive comments and mark-ups in the document.

COMMENT: I have dealt with all comments in the text (which I attach).

1) The level of awareness ('consciousness') of speakers' own variation cannot be gathered from the data as presented here. We simply do not know whether speakers realise they are using different variants or not unless you have explicit information on that. In fact, as soon as you ask informants for 'other words for the same thing', there must be some level of awareness of variation.

 

ANSWER: I added new text to the article. We consider that Basque speakers have the same or similar awareness of variation as speakers of other languages, despite the fact that Basque is said to have greater variety than neighbouring languages. That said, the data analysed in this article are exclusively data produced by the speakers without any motivation or conditioning by the fieldworker, apart from what may be conditioned or motivated unintentionally in a directed but apparently informal and casual conversation. Data expressly motivated (asking for other possible words...), conditioned or suggested by the fieldworker are not the object of analysis in this article; they have been expressly rejected. Only linguistic variation emanating from the speaker's spontaneous locution has been taken into account, which have not been the subject of commentary either by the speaker or by the fieldworker. For this reason, we argue that these are non-conscious data, although we understand that there may be doubts about this in people who do not have first-hand knowledge of the material collected. And because we understand that when a speaker, with no linguistic training and practically illiterate in his language, talks about a subject, he does not worry about the how but about the what, approaching the concept in the best possible way and not worrying about the way it is expressed. In fact, it is a generation that has never worried about the how in its own language, since it is illiterate, not literate in its own language, although it is literate in Spanish or French.

2) In the same vein, more detailed information on the way intraindividual variation was elicited and produced by the speakers would be needed. For example, it makes a huge difference whether a certain lexical variant was first used spontaneously by the speaker and then he/she adds "but my grandfather said XY". Clearly, then, variation is in reception rather than production and, importantly, metalinguistic comments of the latter kind indicate awareness of variation and typically give some information on constraints of variation.

ANSWER: It is true that awareness of variation appears when, once the responses have been elicited, the fieldworker wants to investigate the status of lexical variants in the locality, as in example (15). It is then that speakers realise that they use two different words for the same concept. In case they realised it of their own accord it would not be an unconscious variation and should not be treated, as it has been done in this article. Cases where the informant did not produce the word or variant under analysis on his or her own initiative are not dealt with. Once the word has been produced, it is possible that in some cases and upon questioning by the fieldworker, he/she may specify that his/her parents used this variant or word exclusively, but that in any case he/she or they also use it together with other word(s).

3) In order to be able to argue that the kind of variation encountered is "non-conditioned" as suggested by the title and numerous times in the text, you would have to explicitly and convincingly rule out any extralinguistic AND linguistic factors potentially conditioning variation. It is true that the basic situation and the communicative partners do not change during the interview but the immediate pragmatics and linguistic contexts can and do change nonetheless! There are many instances where some linguistic conditioning may be at work - apart from social/situational constraints which may not be directly applicable but perhaps mentioned in metalinguistic observations.

ANSWER: In agreement, a paragraph has been added to each example with the possible pragmatic and linguistic contexts that may influence the change.

4) In my view, there is a lack of in-depth data analysis. In section 3, you give a basic quantitative overview and then a number of examples. In my view, a more extensive and thorough analysis of results is necessary. This could be quantitative or qualitative in kind, but should set out to answer specific research questions (apart from demonstrating that there IS intra-individual variation).

ANSWER: New text was added to the contribution

5) Also, more background information on the Basque language situation AND the specific variants in the examples would be necessary (e.g., what is their position in the diaglossic spectrum)?

ANSWER: I added more text in the article. As far as Basque background information is concerned, in phonology and pronunciation there is no specific norm for standard Basque; in morphology, syntax and lexicon the standard forms are codified (Gramatika batzordea, 2021, Euskal Gramatika, 2 T., Bibao: Euskaltzaindia. https://www.euskaltzaindia.eus/hizkuntza-baliabideak/gramatika. Euskaltzaindia, 2012, Euskaltzaindiaren hiztegia (the Dictionary of the Academy of the Basque language). https://www.euskaltzaindia.eus/index.php?option=com_hiztegianbilatu&task=hasiera&Itemid=1693&lang=eu .

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

See PDF.

Comments for author File: Comments.pdf

Comments on the Quality of English Language

Author Response

1. Theoretical embedding: The review of previous approaches to IAV in section 1 is rather superficial.... distinctions between non-conditioned IAV, conditioned IAV, and functionalized IAV (see Ulbrich & Werth 2021)... The authors should also go into more detail on how to operationalize there concepts (since this is no trivial matter!). Additionally... presence of “free variation” in language (see.. Bülow & Pfenninger eds (2021): intra-speaker variation across time and space – Sociolinguistics Meets Psycholinguistics, Special Issue in Linguistics Vanguar 7(2). The article clarify what new insights or perspectives it contributes to this body of research. Also, the authors should provide clear research questions

ANSWER: OK. I have introduced the required information in the article: Indeed, the various approaches to IAV have analysed different types or domains. A good approach to the subject is the article by Ulbrich & Werth (2021). These authors have distinguished four types: unconditioned, conditioned, functionalised and mandatory form (Idem::7-11). These authors define unconditioned IAV as variation that ‘cannot be attributed to social, situational or psychological factors considered extra-linguistic, nor by factors inherent to the linguistic system such as phonotactics, morphological conditions or syntactic constraints.’ (Idem:7). We fully agree with Ulbrich and Werth when they state that ‘We claim that not every observable IAV has to be functionalized or conditioned’ (2021:8). As far as the relationship between IAV and IEV is concerned, it must be argued that the linguistic inventory of IAV must inevitably be integrated into that of IEV. The IAV cannot have any feature that does not belong to the IEV. But not the other way around; that is, there can be IEV features in the language community that do not appear in the IAV of all the members of that community. At the same time, not all members of a community have the same variation, neither qualitatively nor quantitatively (some members have more variation than others). This being so, one of the questions that could be asked is the following: what typology should the members who produce the greatest intra-speaker variation have? Presumably it would not be those who use more standardised forms and are more educated. It may be related to the evidence from variationist sociolinguistics that the upper strata of society are not the most conducive to variation.

2. Methodology: atlas survey is not ideal data to investigate IAV.... Questionnaire are never natural speech data. The genre 'interview' itself is an unnatural context of language use... The authors should reflect more thoroughly on how these king of data give insight into IAV.

ANSWER: The Basque atlas surveys have not been of the question-answer, question-answer style until the end of the questionnaire. In this it differed greatly from the atlas tradition. It is clear that the atlas fieldworkers were provided with a questionnaire; but it was not a questionnaire in the usual sense. Apart from the name of the object, the questionnaire asked what it was used for, how it was made, etc. in such a way that the informant had to pronounce more than once the word that was being sought, giving the option of variants during the course of the conversation. The survey sheets addressed how to collect information by conducting a conversation that resembled a casual, informal conversation, in which the interviewer had a notebook on the table and did not write down anything during the interview: everything was recorded, from the beginning to the end. Conversation with even jokes and all kinds of digressions, to make the informant feel comfortable, telling things about his or her life, even private ones. This type of survey-conversation can be ideal, if taken without haste, to produce intra-speaker variation. It is not easy, as others have said, to pick up this kind of variation, it requires a lot of patience and careful direction of the conversation; therefore, we understand as accurate that such variation is difficult to verify (Ulbrich & Werth 2021:8). We understand this context to be appropriate for achieving unconditioned and non-conscious IAV variation, provided that the situational context is the same and the linguistic context is appropriate.

3. Results: … qualitative... In my opinion, incorporating additional quantitative data could provide a more comprehensive picture. For instance, quantifying the relative frequency of IAV in the data, across different linguistics subsystems etc. would benefit the analysis. Regarding the qualitative findings, the authors present different linguistic examples but do not systematically contextualize them. For instance, they could explain the specific contexts in which there variants were used, information on how the variants were elicited, or whether there were any metalinguistic comments from the interviewees.

ANSWER: It has not been possible to analyse all cases of IAV, so a quantitative analysis of their frequency is not possible. But it is worth noting that the context in all of them was a conversation (as specified above) in which the fieldworker directed the conversation in such a way that the informant pronounced the voice or expression on more than one occasion. The utterance of the same word or expression on more than one occasion led to the variants without anyone being aware of it at the time of the survey, although when the researcher transcribed the recording he noticed the variants and wrote them down in the answer booklet which he was obliged to fill in. In general concepts, such as example (11), they may have been uttered in different questions and even on different days.

4. Discussion: In my opinion, this conclusion is not entirely convincing... stylistic variation is expected to occur independently of the addressee (for example, through conversational code-switching, among other factors). The claim that the observed IAV is largely unconscious appers to be speculative....

ANSWER: The conditions described above seem ideal for intra-speaker variation to be unconditioned and unconscious. There may be some hidden factors in some examples that distort the analysis, but in most cases it can be said, at some risk, that these are true cases of unconditioned and unconscious intra-speaker variation. My starting point is the actual data, so it is an empirical work that tries to avoid speculation, even if it may lack specificity in the interpretation of the data or if some interpretations are not sufficiently supported.

5. Regarding the 'framework to analyze intra-speaker variation', I do not see what is novel about these ideas. Similar concepts have been discussed since the 19h c.(...). The authors should either clarifly what is innovative about their approach or refer back to there research traditions.

ANSWER: I agree with this comment. The novelty of the contribution is to present a linguistic variation that is not subject to any motivation or condition and unconscious, not only on a theoretical but also on a practical level, in line with the inherent variation that has been promulgated throughout the history of linguistics and which was focused on in the early days of variationist sociolinguistics. To this end, data are provided to corroborate such variation. I consider my contribution as a small step forward in the defence of the IAV.

For side comments see the corrected version.

Author Response File: Author Response.pdf

Round 2

Reviewer 2 Report

Comments and Suggestions for Authors

Thank you for addressing almost all of the issues raised in my review. In summary, most of these issues have been resolved. A few minor issues and typos remain, which I have marked in the document. 

Comments for author File: Comments.pdf

Comments on the Quality of English Language

The English is fine except for a few minor errors or typos (marked in the document). 

Author Response

Dear reviewer,

All your comments have been accepted.

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

The authors have addressed most of the comments and suggestions in their revision. For those not incorporated, they have provided valid justifications. In its current form, I find the paper appropriate for publication.

Author Response

Dear reviewer,

Thank you for your comments.