The Indeterminacy of Social Meaning Linked to ‘Mexico’ and ‘Texas’ Spanish: Examining Monoglossic Language Ideologies among Heritage and L2 Spanish Listeners

Round 1

Reviewer 1 Report

This manuscript explores a topic that is of interest both to sociolinguists and to scholars of Spanish language education (including Spanish heritage language education) and uses an innovate methodological approach to do so. There is significant interest in the role of attitudes and ideologies in the perception of Spanish and Spanish speakers in the United States. Thus the study reported in the manuscript has the potential to be of interest for many readers. As much as I wanted to like the manuscript, however, I struggled to understand the motivation for the research design or the contribution. 

One immediate (very easily fixable) issue was the confusing use of terms without sufficient explanation or examples. Specifically, “perception” and “evaluation” do not seem to be used consistently throughout the paper (or defined), and it would be helpful to clarify what is meant by “linguistic perception” (vs “linguistic evaluation”) and “social perception” vs “social evaluation”. For example, on the second paragraph of page 2 matched guise studies are referred to as “social perception experiment[s]” and “sociolinguistic perception studies”. In this paragraph, the authors explain that changing a single linguistic variable can affect “listeners’ evaluations” and “social evaluations;” it would be very helpful to explain what is meant by this and to give concrete examples of such evaluations. In this paragraph, the sentence “Many of these studies have shown that there is a multitude of social meaning attached to the linguistic variants based on speaker and listener characteristics” is unclear. I can’t tell if the point is that the listeners/participants assume that speakers have particular characteristics based on the linguistic variants they use, or that particular variants carry a lot of social meaning. Either way, it would help to have an example.) The potential for confusion about “perception” is increased by the existence of “perceptual dialectology” studies that also looked at the interaction of the linguistic and the social, but without using any recordings. It definitely seems like Alfaraz’s and Montes-Alcala’s perceptual dialectology research should be included, because it focuses on attitudes/perceptions of different varieties of Spanish by people in the US.

Related to the need for clarifying “perception”, in the discussion of studies in which participants were led to believe that a recorded speaker was of a specific nationality, it is a little confusing to call this “perceived nationality” when the focus of the study was on how participants ‘perceived/heard’ linguistic information. Isn’t this a case social information affecting linguistic perception? I understand that “perceived nationality” is probably meant to make clear that these were not ‘actual’ nationalities but in the context of the (good and sound) argument that 1) social information influences linguistic perception and 2) linguistic information influences social perception, it seems like it would be clearer not to have “perception” on both sides of the causal relationship. It might be better to say “information about nationality” or “implied nationality” or “reported nationality”. 

My biggest concern with the manuscript is that although the authors do a good job arguing that the relationship between language and the social is two-way, there is an URGENT need to explain and motivate the study of how “social information impacts social perception.” Following the argument about the bidirectional relationship of linguistic info and social info, it is not clear what this even means (in introduction and elsewhere), and it sounds like a tautology. In what are the only examples provided, the authors discuss Rubin (1992) and Gutierrez and Amengual (2016), but both those studies looked at whether social information (visual cues regarding ethnicity or race) affected judgments of accentedness (Rubin) and comprehension, and these seem like LINGUISTIC evaluations rather than “social perceptions”. In the Gutierrez and Amengual study, results showed that participants used phenotype or other visual cues to ascribe ethnolinguistic identity, and that ‘looking Hispanic’ was tied to expectations of Spanish-speaking. It’s hard to see how that work motivates this study. If we take at face value the stated goal to study whether social information impacts social perception, what is the motivation for including the recordings (and for making the ASSUMPTION that participants are ‘really’ comparing based on 2 varieties of Spanish)? At the most basic level, isn’t this just a question of whether participants judge (stereotype) persons differently based on whether they are identified as Mexican or Mexican American? Or is the idea that the social information impacts linguistic perception which in turn impacts social evaluation? If this chain of influence is the point it should be explained, but it seems to me that rather than a 3^rd area of research (#1 = linguistic impacts social evaluation; #2 social impacts linguistic perception; #3 social info impacts social perception), the most interesting results of this study are related to #1 and #2.

Another serious concern about the motivation and research design is the idea that origin (ie “from Mexico” vs “from Texas”) is simply a proxy for monolingual vs bilingual status. In the case of the open-ended language evaluation, this might be more justifiable (e.g., participants ‘hearing’ people from Mexico as speaking faster or those from Texas as “stumbling”), but for social evaluation, but in the case of the other evaluations (socioeconomic status, education, friendliness, etc.), why would we assume these are based on language, rather than national stereotypes, and/or awareness of different patterns of racialization, or ? The title and purported focus of the paper seem overstated, especially given the lack of bilingual or contact features in the recorded stimuli or open-ended questions; it’s not clear that this paper actually investigated monoglossic ideologies. 

As for the results and discussion, I am not convinced by the authors’ explanation of “perceived monolingual speakers” being rated higher on “ability to help with one’s Spanish” (the sole main effect of the study); the authors argue that this reflects a monoglossic ideology and the idea that monolinguals are better at teaching or explaining than bilinguals. Isn’t it equally (or more) likely that participants are aware that people from Mexico (regardless of whether they are mono- bi- or multilingual) will have received formal education in Spanish, but Spanish-speakers from Texas are likely to have had most of their formal education only in English? In other words, as noted, “from Texas” or “from Mexico” is a proxy for many different variables, including whether or not one has had formal education in Spanish, which might be considered valuable for the ability to teach or explain Spanish to the participants. Also, I would have liked to see more discussion or interpretation of the findings. For example, why did HL speakers (but not L2 speakers) perceive Texas as being of lower socioeconomic status? (If the authors have no explanation, they might say so). Similarly, how do the authors account for the gender effects?

To be honest, I’m not certain whether these issues can be satisfactorily resolved with the existing data. Assuming that they can, or of the authors continue to work on the topic, there are also several problems and/or areas for improvement in the literature review sections. 

-       The definition/discussion of monoglossic ideologies needs work. One can certainly argue that monoglossic ideologies are related to the one nation one language ideology, but these are not the same thing. Also, monoglossic ideologies are not only about the evaluation of monolingual vs bilingual varieties but all “non-standard” varieties.

-       The issue of linguistic insecurity is an interesting one that is raised in the paper. On page 5, the authors suggest that linguistic insecurity is the main cause of language shift. However, most researchers agree that the lack of institutional, societal and educational support is the primary cause, and perhaps they really mean that language shift contributes to linguistic insecurity. In any case, it would be useful to clarify the difference between attitudes towards bilingual speech (monoglossic ideologies) and insecurities about language loss/shift. 

-       The authors seem to emphasize the agentive use of language in the creating social meaning but then their study focuses on perception. On page 2 they acknowledge that most studies look at discourse in interaction and then go on to talk about matched guise studies. I would have liked to see this fleshed out with a discussion/argument of why perception matters, or what its link is to the use of language (variation) in interaction or production. This is important not only for motivating the study but also for making it ‘fit’ within the special issue.

-       The discussion of the 3 ‘waves’ could be tightened. The authors vacillate regarding whether the different weaves looked at social meaning or not. But in any case, I wasn’t sure it is necessary or valuable to get into a discussion of all 3 waves. Couldn’t this section just talk about current understanding? OR, if the authors choose to keep this in, the importance/relevance should be made clear (for example, speakers/listeners associate particular ways of speaking with specific groups, which will be relevant for discussion of matched guise later)

-       The organization of the lit review could be improved. As is, the “background” section 2.1 covers a lot of the key points with a synthetic discussion of the issues, and then a traditional study-by-study lit review is given in section 2.3.  It might also help not to have these sections split by the current section 2.2 on language ideologies. My suggestion would be to remove PERCEPTION from current section 2.1 and possibly combine it with section current section 2.2 ideologies earlier, focusing on the broad relationship between the social and the linguistic (which are linked via ideologies) and then get into the discussion of how this has been examined quantitatively in PERCEPTION (which could be either one or two sections). 

-       Although the authors clearly situate the paper within the quantitative approach, they may also wish to include some qualitative research on the role of social information on linguistic perception, such as work by Inoue (2006) or Rosa and Flores.

Author Response

Please see attached.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments for author File: Comments.pdf

Author Response

Please see attached. 

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

The authors have done a very nice job responding to comments and editing the manuscript. It is now much clearer, and greatly improved. Congratulations!

The clarifications and definitions of the different constructs added to the first paragraph are a huge improvement, though I still have some suggestions.

1.     The authors now give this definition of social perception:  “SOCIAL PERCEPTION is defined as the social characteristics listeners’ attribute to different speakers based on the linguistic input they hear such as perceived educational level, friendliness, respectfulness, etc.:   If this study looks at the role of SOCIAL information on social perception, shouldn’t “based on the linguistic input they hear” be removed?

2.     They go on to say that “fewer studies examining the role of social information on social perception and/or linguistic evaluations. That is to say, without manipulating the linguistic input, can the mere presence of social information affect how listeners perceive one’s speech?”  I have two concerns here. First, according to your previous distinction between “linguistic perception” and “linguistic evaluation”, shouldn’t this be “That is to say, without manipulating the linguistic input, can the mere presence of social information affect how listeners evaluate one’s speech?” Second, although this explains “the role of social information on linguistic evaluations” it would be very helpful to include an explanation of what is meant by the other possibility (namely, “the role of social information on social perception”).

The following paragraph (and the results section), now states that the current study looks at the interaction of linguistic and social information. On the one hand, this seems like it addressed my (primary) concern that looking at the impact of social information on social perception is basically a study of stereotypes, and it seemed like an improvement. But as far as I can tell, the study didn’t really investigate the interaction of these two kinds of information (since the linguistic information was not a variable). I still quite get my head around the motivation for the research design, and it would help to include more explanation. If it is about the impact of social information on social perception, the authors should explain why they include any linguistic input. (I guess this was to make the experiment seem more ‘natural’ but it means that there was an additional mediating factor).  The qualitative component seems like a cleaner case of social information  linguistic evaluation.  

It would be helpful to have a short summary paragraph at the end of the section with quantitative findings (presented in the table), before discussing the word clouds.

I am not completely convinced by the design and assumptions, but the  limitations are now mentioned. The article is well-written and well-argued, and it deserves to be published.

Author Response

Please see attachment.

Author Response File: Author Response.pdf