|
Post by Admin on Nov 3, 2015 15:29:02 GMT
Welcome to the Language and New Media e-Seminar! Over the next two weeks, we will be reading Camilla Vasquez’s recent paper: ‘Right now versus back then: Recency and remoteness as discursive resources in online reviews’ Discourse, Context & Media, (available online 9 June 2015). You can download a copy of this article for free from Discourse, Context & Media: www.sciencedirect.com/science/article/pii/S2211695815000252We'll be using this forum as the place to discuss how the ideas and methods in the paper might apply to the research interests of the people in the group, and to think about how the paper raises questions about the future directions of work in language and the media. Please post your responses as replies to this thread. To get us started in sharing ideas, please tell us how the paper related to your own research. Do you think the emphasis on 'recency' or 'remoteness' was more apparent?
|
|
Erika
New Member
Posts: 2
|
Post by Erika on Nov 3, 2015 15:47:20 GMT
Hi, it's great to be here and I am looking forward to the discussion
|
|
|
Post by Caroline on Nov 3, 2015 17:46:49 GMT
Sounds great!
|
|
|
Post by tereza on Nov 12, 2015 17:29:38 GMT
All set and ready for our virtual natter between Nov 23rd and Dec 8th. Looking forward to!
|
|
|
Post by Caroline on Nov 23, 2015 17:49:45 GMT
I enjoyed reading this article and think its focus on 'time' and social media is very timely (pun intended!). It also resonates with some thinking that Agnieszka (Lyons) and I are doing around the social media data collected as part of a large ethnographic project (http://www.birmingham.ac.uk/generic/tlang/index.aspx). The social media data comprises SMS, WhatsApp and Viber messages sent by people who run small businesses in the UK, and we are interested in how virtual messages interact with physical contexts. In particular, what seems interesting in our data in the effect of social media on space and time or, rather, how people can exploit space and time as a communicative resource in mediated interactions. I like the way that Camilla uses precise linguistic analysis to show how people can create and exploit different time-spaces, and how these act as persuasive devices rather straightforward deictic references.
My question for Camilla would be whether you have access to actual time stamps and/or any idea as to the timeframe in which people read and respond to reviews; and, if you do, any way of comparing the timeliness of review sites to other sites such as Facebook? Do people tend to read reviews that are posted in the more ‘remote’ past and, if so, do these people respond differently to those who read reviews very ‘recently’ posted? Or is the assumption that these reviews will be read immediately or not at all?
|
|
|
Post by Admin on Nov 24, 2015 15:44:01 GMT
When I read the article about 'remoteness' and 'recency', I was really glad to see Camilla drawing attention to the variety of ways in which time was represented in the different kinds of reviews. The research I am working on at the minute examines a dataset of several hundred YouTube videos about Oscar Pistorius. References to time appear in the 'Description' section of the videos, but also appear most prominently in the videos that are categorised as 'News' (rather than 'Sport', for example). In my data, these references are lexically represented as adverbial phrases that contextualise the breaking news of the reports that were uploaded as video content, so they have become 'frozen in time' as constructions of 'recency', which I interpret as supporting the news values in the YouTube video. It would be interesting to collect comparative data from other YouTube genres to see if the references to time are also included in other types of Description too.
YouTube also seems to be frustrating in terms of how it represents time stamps for comments as they become more remote. They are simply captured as 'one year ago' or similar, so it's hard to do precise analysis with that kind of meta-data.
I'd be interested to find out more about how others in the group analyse temporal reference (linguistic or automated) in social media data too.
|
|
|
Post by richard on Nov 26, 2015 14:18:37 GMT
I found the article very interesting and particularly liked the way that it argued against the eternal present of some descriptions of social media. I thought the differences between the way time is handled in different sites a useful reminder that conceptualisations of time can be very different. The ways that time is treated in the restaurant reviews do suggest that the reviewers think of time differently from the recipe reviewers.
The signals of time orientation were quite varied and I wondered if this kind of analysis could be automated. The use of the past simple to talk about the immediate and remote past indicated that the label of past simple is very misleading but also that data need to be looked at closely to be sure of how they should be classified. The analysis of the present perfect make me consider whether there was a intermediate time frame which linked the past to the present.
|
|
|
Post by Camilla Vasquez on Nov 28, 2015 12:20:03 GMT
When I read the article about 'remoteness' and 'recency', I was really glad to see Camilla drawing attention to the variety of ways in which time was represented in the different kinds of reviews. The research I am working on at the minute examines a dataset of several hundred YouTube videos about Oscar Pistorius. References to time appear in the 'Description' section of the videos, but also appear most prominently in the videos that are categorised as 'News' (rather than 'Sport', for example). In my data, these references are lexically represented as adverbial phrases that contextualise the breaking news of the reports that were uploaded as video content, so they have become 'frozen in time' as constructions of 'recency', which I interpret as supporting the news values in the YouTube video. It would be interesting to collect comparative data from other YouTube genres to see if the references to time are also included in other types of Description too. YouTube also seems to be frustrating in terms of how it represents time stamps for comments as they become more remote. They are simply captured as 'one year ago' or similar, so it's hard to do precise analysis with that kind of meta-data. I'd be interested to find out more about how others in the group analyse temporal reference (linguistic or automated) in social media data too. Ruth, I love your description of “constructions of recency as frozen in time” in your YouTube data – what a great observation! Otherwise, your comment hits on what was, for me, one of the most surprising findings of my study – which is that time is represented in such a range of different ways in these online data. I must confess that when I first embarked on this project, I didn’t imagine there would be so many different linguistic resources being used to refer to time in reviews. Similarly, I also thought that the classification of “recent vs. not” would be more straightforward than it was – and I hope that where I’ve discussed my methodological decision-making in the article I’ve done justice to some of the fuzziness involved in this process (as was the case of present perfect, that Richard points out in his comment below).
|
|
|
Post by Camilla Vasquez on Nov 28, 2015 12:21:23 GMT
I found the article very interesting and particularly liked the way that it argued against the eternal present of some descriptions of social media. I thought the differences between the way time is handled in different sites a useful reminder that conceptualisations of time can be very different. The ways that time is treated in the restaurant reviews do suggest that the reviewers think of time differently from the recipe reviewers. The signals of time orientation were quite varied and I wondered if this kind of analysis could be automated. The use of the past simple to talk about the immediate and remote past indicated that the label of past simple is very misleading but also that data need to be looked at closely to be sure of how they should be classified. The analysis of the present perfect make me consider whether there was a intermediate time frame which linked the past to the present. Thanks so much for your insights, Richard. I appreciate your inquiry about automating the process. Certainly a POS tagger could help identify features like tense and aspect, which may offer a useful starting point. However, for some of the other linguistic features used to refer to time (I am thinking here, for example, about prepositional phrases and adverbial clauses), the identification of a temporal reference is more a matter of semantics than of grammar, so I don’t actually know if/how that part could be automated. (Perhaps someone who is more experienced with these sorts of computational tools than I am might be able to offer further suggestions about this?) Otherwise, your point that the “data need to be looked at closely to be sure of how they should be classified” is definitely well-taken. Even if one were to use an automated procedure, a close and careful examination of the data by the researcher(s) would also be necessary to ensure the accuracy of classifications – and to make sure that nothing potentially relevant gets overlooked.
|
|
|
Post by Camilla Vasquez on Nov 28, 2015 12:22:54 GMT
I enjoyed reading this article and think its focus on 'time' and social media is very timely (pun intended!). It also resonates with some thinking that Agnieszka (Lyons) and I are doing around the social media data collected as part of a large ethnographic project (http://www.birmingham.ac.uk/generic/tlang/index.aspx). The social media data comprises SMS, WhatsApp and Viber messages sent by people who run small businesses in the UK, and we are interested in how virtual messages interact with physical contexts. In particular, what seems interesting in our data in the effect of social media on space and time or, rather, how people can exploit space and time as a communicative resource in mediated interactions. I like the way that Camilla uses precise linguistic analysis to show how people can create and exploit different time-spaces, and how these act as persuasive devices rather straightforward deictic references. My question for Camilla would be whether you have access to actual time stamps and/or any idea as to the timeframe in which people read and respond to reviews; and, if you do, any way of comparing the timeliness of review sites to other sites such as Facebook? Do people tend to read reviews that are posted in the more ‘remote’ past and, if so, do these people respond differently to those who read reviews very ‘recently’ posted? Or is the assumption that these reviews will be read immediately or not at all? Caroline, your and Agnieszka’s work sounds really interesting – I’d love to know more about how users of those applications that you mention exploit temporal references, and the ways in which they may be similar to/different from those in reviews. Thank you for your great questions as well! To answer your first question about timestamps: it totally varies by site. On most of the sites (Yelp, Epicurious, Amazon) the timestamp does appear with each review as the date that the review was posted. TripAdvisor (which seems to constantly be making changes to its interface) used to include the date that the review was posted, but now does the more relative “posted 2 weeks ago” thing that Ruth mentioned above with respect to YouTube comments. And Netflix (the site that included the least amount of reviewer info & meta-data) did not include any kind of a timestamp with reviews. However, things get more complicated when we take into consideration each site’s default display mechanism, which also varies. In the most straightforward cases (e.g., Epicurious, TripAdvisor) reviews are displayed in order of recency of posting, which we can observe by simply looking at the timestamps and the reverse-chronological order in which consecutive reviews appear. But on other sites, reviews are displayed according to some algorithm that additionally factors in the number of helpfulness votes that each review has received. So an older review with more “helpfulness” votes might actually appear before one that has been posted more recently. This happens on both Amazon and Yelp, and it creates kind of a Catch-22 situation: the more helpfulness votes a review gets, the more prominently it gets displayed --> the more prominently it gets displayed, the more views it gets --> the more views it gets, the more likely it is to get voted on as “helpful” by others… and to continue to maintain its position. This cycle becomes evident when we look at products that have been the target of parody reviews on Amazon, for example, many of which have thousands of reviews – with helpfulness votes numbering in the TENS of thousands. The first review that appears for one of these products (the Hutzler Banana Slicer, for instance) may be from 2011 (with 30,000+ helpfulness votes), while the most recent review (with 1 helpfulness vote) was posted just yesterday and is buried somewhere on page 3,000 of results. That said, these sites do enable users to adjust the sorting process by selecting “newest first,” or by selecting a specific star rating (or, on TripAdvisor, you can now event sort by the “season” in which a hotel review was posted!) … but it would be interesting to know how many users actually take advantage of this feature, rather than just relying on the site’s default sorting mechanism. Answering this question would require taking a user-based approach. As for your question about how these processes compare to the order of posting/timeliness of responding on Facebook … it would be fascinating to hear what others who have explored time on Facebook have found in this regard!
|
|
|
Post by agnieszka on Dec 2, 2015 19:13:19 GMT
As I was reading the article, it occurred to me that in some cases the expression of time, be it recency or remoteness, seems to serve the construction of the reviewer's image. We have an insight into their lives and can imagine what type of persona they present through their reference to time (a person who spends their lunch time eating at a restaurant by themselves, a loyal user of loose tea, blender expert with years of experience, someone subscribing to their mother's traditional cooking tips, etc).
There is also the question of how long you have to have owned a product to justify a (positive) review and how the length of time then corresponds to how the person presents themselves, e.g., a keen and extensive user of a brand new product (still excited to own it) or an expert with no mention of frequency or intensity of use.
This in turn brings the question of the process of production of these texts. I would be interested to find out why people decide to post their reviews at those particular times. Certain websites will prompt you to write a review after a few days/weeks/months (?). I wondered if this affects the temporal reference in any way...
|
|
|
Post by Caroline on Dec 3, 2015 11:15:02 GMT
As I was reading the article, it occurred to me that in some cases the expression of time, be it recency or remoteness, seems to serve the construction of the reviewer's image. We have an insight into their lives and can imagine what type of persona they present through their reference to time (a person who spends their lunch time eating at a restaurant by themselves, a loyal user of loose tea, blender expert with years of experience, someone subscribing to their mother's traditional cooking tips, etc). There is also the question of how long you have to have owned a product to justify a (positive) review and how the length of time then corresponds to how the person presents themselves, e.g., a keen and extensive user of a brand new product (still excited to own it) or an expert with no mention of frequency or intensity of use. Agnieszka, I'm reminded of the chapter Camilla wrote for my and Philip's book ('The Language of Social Media' Palgrave 2014), where she talks about how reviewers do exactly what you've identified - draw on relevant aspects of their background, experiences and expertise in order to persuade others that you are credible. Other reviewers then evaluate the review based on the extent to which they align with the reviewer - are they the same kind of person, etc. The chapter is '"Usually not one to complain but ...": constructing identities in user-generated online reviews'.
|
|
|
Post by tereza on Dec 3, 2015 12:10:29 GMT
I thoroughly enjoyed reading the article, as well as the posts above. Camilla's paper offers important insights into the relative meaning of temporal indexicals and variation in temporal perspectives across different websites. What I found particularly interesting is the finding about variation in temporal perspectives even within a single website, leading to explorations about how time is used as a communicative resource to display e.g. reviewer's expertise, quality of reviewed product, etc.
Reflecting upon how time has 'featured' in my work, both linguistic and automated time (e.g. time stamps) in the texting data of my PhD (collected way back in 2003) were key resources for mundane interactional tasks (e.g. micro-coordination), as well as for displaying local social identities (e.g. constructing one another as BFF). While trying to make sense of such private interactions, ethnographic insights as to how my participants' (offline) life was organised (in terms of the 'timing' of their daily activities) were very useful, in order to understand, for instance, how/why a particular reply was deemed 'delayed' or not. While reading the posts above, I gathered that Camilla and colleagues have been pointing to further layers according to which social media activities get contextualised: not only in terms of how they are embedded in other types of participant activities (e.g. reading reviews, visiting restaurants, etc), but also in terms of how such activities are mediated by the system interface (e.g. algorithms) and the user's interaction with such aspects of the system (e.g. viewing, voting, liking, etc). I sense that this may be relevant when researching temporality in less public mediated interaction (compared to online reviews), such as, perhaps, posts (replies, etc) on Facebook News Feed.
My question to Camilla is whether a similar approach can be undertaken to the exploration of spatial references in online reviews. I understand that this is way beyond the aims of the paper. But echoing Caroline's phrase of 'time-spaces' above and thinking that it might be more relevant for certain review sites (e.g. for hotels or restaurants) than others, I am wondering whether spatial 'proximity' and 'distance' have also featured as communicative resources to achieve various interactional tasks in these online environments.
|
|
|
Post by agnieszka on Dec 3, 2015 17:04:32 GMT
As I was reading the article, it occurred to me that in some cases the expression of time, be it recency or remoteness, seems to serve the construction of the reviewer's image. We have an insight into their lives and can imagine what type of persona they present through their reference to time (a person who spends their lunch time eating at a restaurant by themselves, a loyal user of loose tea, blender expert with years of experience, someone subscribing to their mother's traditional cooking tips, etc). There is also the question of how long you have to have owned a product to justify a (positive) review and how the length of time then corresponds to how the person presents themselves, e.g., a keen and extensive user of a brand new product (still excited to own it) or an expert with no mention of frequency or intensity of use. Agnieszka, I'm reminded of the chapter Camilla wrote for my and Philip's book ('The Language of Social Media' Palgrave 2014), where she talks about how reviewers do exactly what you've identified - draw on relevant aspects of their background, experiences and expertise in order to persuade others that you are credible. Other reviewers then evaluate the review based on the extent to which they align with the reviewer - are they the same kind of person, etc. The chapter is '"Usually not one to complain but ...": constructing identities in user-generated online reviews'. Yes! Thank you for the reminder. I guess what I would be interested in is a close-up on the time aspect in the construction of self-presentation. The same, in fact, would apply to space, which Tereza mentioned and which resonates with my deictic shift focus in my text-message research.
|
|
|
Post by Caroline on Dec 4, 2015 15:32:12 GMT
Yes! Thank you for the reminder. I guess what I would be interested in is a close-up on the time aspect in the construction of self-presentation. The same, in fact, would apply to space, which Tereza mentioned and which resonates with my deictic shift focus in my text-message research. Gotcha! Yes, the time aspect would presumably tie in with creating a status as expert and experienced, among other things; I bet it would also tie in with identity constructions such as being very busy and on the go, that's a prevalent identity stance in my experience! And yes, I don't think that time can be easily isolated from space. The constructed time and space create a frame within which the interaction - the review? - can be understood and evaluated.
|
|