Tag Archives: Médialogues

Social Networks and Microblogging

Microblogging (Laconica, Twitter, etc.) is still a hot topic. For instance, during the past few episodes of This Week in Tech, comments were made about the preponderance of Twitter as a discussion theme: microblogging is so prominent on that show that some people complain that there’s too much talk about Twitter. Given the centrality of Leo Laporte’s podcast in geek culture (among Anglos, at least), such comments are significant.

The context for the latest comments about TWiT coverage of Twitter had to do with Twitter’s financials: during this financial crisis, Twitter is given funding without even asking for it. While it may seem surprising at first, given the fact that Twitter hasn’t publicized a business plan and doesn’t appear to be profitable at this time, 

Along with social networking, microblogging is even discussed in mainstream media. For instance, Médialogues (a media critique on Swiss national radio) recently had a segment about both Facebook and Twitter. Just yesterday, Comedy Central’s The Daily Show with Jon Stewart made fun of compulsive twittering and mainstream media coverage of Twitter (original, Canadian access).

Clearly, microblogging is getting some mindshare.

What the future holds for microblogging is clearly uncertain. Anything can happen. My guess is that microblogging will remain important for a while (at least a few years) but that it will transform itself rather radically. Chances are that other platforms will have microblogging features (something Facebook can do with status updates and something Automattic has been trying to do with some WordPress themes). In these troubled times, Montreal startup Identi.ca received some funding to continue developing its open microblogging platform.  Jaiku, bought by Google last year, is going open source, which may be good news for microblogging in general. Twitter itself might maintain its “marketshare” or other players may take over. There’s already a large number of third-party tools and services making use of Twitter, from Mahalo Answers to Remember the Milk, Twistory to TweetDeck.

Together, these all point to the current importance of microblogging and the potential for further development in that sphere. None of this means that microblogging is “The Next Big Thing.” But it’s reasonable to expect that microblogging will continue to grow in use.

(Those who are trying to grok microblogging, Common Craft’s Twitter in Plain English video is among the best-known descriptions of Twitter and it seems like an efficient way to “get the idea.”)

One thing which is rarely mentioned about microblogging is the prominent social structure supporting it. Like “Social Networking Systems” (LinkedIn, Facebook, Ning, MySpace…), microblogging makes it possible for people to “connect” to one another (as contacts/acquaintances/friends). Like blogs, microblogging platforms make it possible to link to somebody else’s material and get notifications for some of these links (a bit like pings and trackbacks). Like blogrolls, microblogging systems allow for lists of “favourite authors.” Unlike Social Networking Systems but similar to blogrolls, microblogging allow for asymmetrical relations, unreciprocated links: if I like somebody’s microblogging updates, I can subscribe to those (by “following” that person) and publicly show my appreciation of that person’s work, regardless of whether or not this microblogger likes my own updates.

There’s something strangely powerful there because it taps the power of social networks while avoiding tricky issues of reciprocity, “confidentiality,” and “intimacy.”

From the end user’s perspective, microblogging contacts may be easier to establish than contacts through Facebook or Orkut. From a social science perspective, microblogging links seem to approximate some of the fluidity found in social networks, without adding much complexity in the description of the relationships. Subscribing to someone’s updates gives me the role of “follower” with regards to that person. Conversely, those I follow receive the role of “following” (“followee” would seem logical, given the common “-er”/”-ee” pattern). The following and follower roles are complementary but each is sufficient by itself as a useful social link.

Typically, a microblogging system like Twitter or Identi.ca qualifies two-way connections as “friendship” while one-way connections could be labelled as “fandom” (if Andrew follows Betty’s updates but Betty doesn’t follow Andrew’s, Andrew is perceived as one of Betty’s “fans”). Profiles on microblogging systems are relatively simple and public, allowing for low-involvement online “presence.” As long as updates are kept public, anybody can connect to anybody else without even needing an introduction. In fact, because microblogging systems send notifications to users when they get new followers (through email and/or SMS), subscribing to someone’s update is often akin to introducing yourself to that person. 

Reciprocating is the object of relatively intense social pressure. A microblogger whose follower:following ratio is far from 1:1 may be regarded as either a snob (follower:following much higher than 1:1) or as something of a microblogging failure (follower:following much lower than 1:1). As in any social context, perceived snobbery may be associated with sophistication but it also carries opprobrium. Perry Belcher  made a video about what he calls “Twitter Snobs” and some French bloggers have elaborated on that concept. (Some are now claiming their right to be Twitter Snobs.) Low follower:following ratios can result from breach of etiquette (for instance, ostentatious self-promotion carried beyond the accepted limit) or even non-human status (many microblogging accounts are associated to “bots” producing automated content).

The result of the pressure for reciprocation is that contacts are reciprocated regardless of personal relations.  Some users even set up ways to automatically follow everyone who follows them. Despite being tricky, these methods escape the personal connection issue. Contrary to Social Networking Systems (and despite the term “friend” used for reciprocated contacts), following someone on a microblogging service implies little in terms of friendship.

One reason I personally find this fascinating is that specifying personal connections has been an important part of the development of social networks online. For instance, long-defunct SixDegrees.com (one of the earliest Social Networking Systems to appear online) required of users that they specified the precise nature of their relationship to users with whom they were connected. Details escape me but I distinctly remember that acquaintances, colleagues, and friends were distinguished. If I remember correctly, only one such personal connection was allowed for any pair of users and this connection had to be confirmed before the two users were linked through the system. Facebook’s method to account for personal connections is somewhat more sophisticated despite the fact that all contacts are labelled as “friends” regardless of the nature of the connection. The uniform use of the term “friend” has been decried by many public commentators of Facebook (including in the United States where “friend” is often applied to any person with whom one is simply on friendly terms).

In this context, the flexibility with which microblogging contacts are made merits consideration: by allowing unidirectional contacts, microblogging platforms may have solved a tricky social network problem. And while the strength of the connection between two microbloggers is left unacknowledged, there are several methods to assess it (for instance through replies and republished updates).

Social contacts are the very basis of social media. In this case, microblogging represents a step towards both simplified and complexified social contacts.

Which leads me to the theme which prompted me to start this blogpost: event-based microblogging.

I posted the following blog entry (in French) about event-based microblogging, back in November.

Microblogue d’événement

I haven’t received any direct feedback on it and the topic seems to have little echoes in the social media sphere.

During the last PodMtl meeting on February 18, I tried to throw my event-based microblogging idea in the ring. This generated a rather lengthy between a friend and myself. (Because I don’t want to put words in this friend’s mouth, who happens to be relatively high-profile, I won’t mention this friend’s name.) This friend voiced several objections to my main idea and I got to think about this basic notion a bit further. At the risk of sounding exceedingly opinionated, I must say that my friend’s objections actually comforted me in the notion that my “event microblog” idea makes a lot of sense.

The basic idea is quite simple: microblogging instances tied to specific events. There are technical issues in terms of hosting and such but I’m mostly thinking about associating microblogs and events.

What I had in mind during the PodMtl discussion has to do with grouping features, which are often requested by Twitter users (including by Perry Belcher who called out Twitter Snobs). And while I do insist on events as a basis for those instances (like groups), some of the same logic applies to specific interests. However, given the time-sensitivity of microblogging, I still think that events are more significant in this context than interests, however defined.

In the PodMtl discussion, I frequently referred to BarCamp-like events (in part because my friend and interlocutor had participated in a number of such events). The same concept applies to any event, including one which is just unfolding (say, assassination of Guinea-Bissau’s president or bombings in Mumbai).

Microblogging users are expected to think about “hashtags,” those textual labels preceded with the ‘#’ symbol which are meant to categorize microblogging updates. But hashtags are problematic on several levels.

  • They require preliminary agreement among multiple microbloggers, a tricky proposition in any social media. “Let’s use #Bissau09. Everybody agrees with that?” It can get ugly and, even if it doesn’t, the process is awkward (especially for new users).
  • Even if agreement has been reached, there might be discrepancies in the way hashtags are typed. “Was it #TwestivalMtl or #TwestivalMontreal, I forgot.”
  • In terms of language economy, it’s unsurprising that the same hashtag would be used for different things. Is “#pcmtl” about Podcamp Montreal, about personal computers in Montreal, about PCM Transcoding Library…?
  • Hashtags are frequently misunderstood by many microbloggers. Just this week, a tweep of mine (a “peep” on Twitter) asked about them after having been on Twitter for months.
  • While there are multiple ways to track hashtags (including through SMS, in some regions), there is no way to further specify the tracked updates (for instance, by user).
  • The distinction between a hashtag and a keyword is too subtle to be really useful. Twitter Search, for instance, lumps the two together.
  • Hashtags take time to type. Even if microbloggers aren’t necessarily typing frantically, the time taken to type all those hashtags seems counterproductive and may even distract microbloggers.
  • Repetitively typing the same string is a very specific kind of task which seems to go against the microblogging ethos, if not the cognitive processes associated with microblogging.
  • The number of character in a hashtag decreases the amount of text in every update. When all you have is 140 characters at a time, the thirteen characters in “#TwestivalMtl” constitute almost 10% of your update.
  • If the same hashtag is used by a large number of people, the visual effect can be that this hashtag is actually dominating the microblogging stream. Since there currently isn’t a way to ignore updates containing a certain hashtag, this effect may even discourage people from using a microblogging service.

There are multiple solutions to these issues, of course. Some of them are surely discussed among developers of microblogging systems. And my notion of event-specific microblogs isn’t geared toward solving these issues. But I do think separate instances make more sense than hashtags, especially in terms of specific events.

My friend’s objections to my event microblogging idea had something to do with visibility. It seems that this friend wants all updates to be visible, regardless of the context. While I don’t disagree with this, I would claim that it would still be useful to “opt out” of certain discussions when people we follow are involved. If I know that Sean is participating in a PHP conference and that most of his updates will be about PHP for a period of time, I would enjoy the possibility to hide PHP-related updates for a specific period of time. The reason I talk about this specific case is simple: a friend of mine has manifested some frustration about the large number of updates made by participants in Podcamp Montreal (myself included). Partly in reaction to this, he stopped following me on Twitter and only resumed following me after Podcamp Montreal had ended. In this case, my friend could have hidden Podcamp Montreal updates and still have received other updates from the same microbloggers.

To a certain extent, event-specific instances are a bit similar to “rooms” in MMORPG and other forms of real-time many-to-many text-based communication such as the nostalgia-inducing Internet Relay Chat. Despite Dave Winer’s strong claim to the contrary (and attempt at defining microblogging away from IRC), a microblogging instance could, in fact, act as a de facto chatroom. When such a structure is needed. Taking advantage of the work done in microblogging over the past year (which seems to have advanced more rapidly than work on chatrooms has, during the past fifteen years). Instead of setting up an IRC channel, a Web-based chatroom, or even a session on MSN Messenger, users could use their microblogging platform of choice and either decide to follow all updates related to a given event or simply not “opt-out” of following those updates (depending on their preferences). Updates related to multiple events are visible simultaneously (which isn’t really the case with IRC or chatrooms) and there could be ways to make event-specific updates more prominent. In fact, there would be easy ways to keep real-time statistics of those updates and get a bird’s eye view of those conversations.

And there’s a point about event-specific microblogging which is likely to both displease “alpha geeks” and convince corporate users: updates about some events could be “protected” in the sense that they would not appear in the public stream in realtime. The simplest case for this could be a company-wide meeting during which backchannel is allowed and even expected “within the walls” of the event. The “nothing should leave this room” attitude seems contradictory to social media in general, but many cases can be made for “confidential microblogging.” Microblogged conversations can easily be archived and these archives could be made public at a later date. Event-specific microblogging allows for some control of the “permeability” of the boundaries surrounding the event. “But why would people use microblogging instead of simply talking to another?,” you ask. Several quick answers: participants aren’t in the same room, vocal communication is mostly single-channel, large groups of people are unlikely to communicate efficiently through oral means only, several things are more efficiently done through writing, written updates are easier to track and archive…

There are many other things I’d like to say about event-based microblogging but this post is already long. There’s one thing I want to explain, which connects back to the social network dimension of microblogging.

Events can be simplistically conceived as social contexts which bring people together. (Yes, duh!) Participants in a given event constitute a “community of experience” regardless of the personal connections between them. They may be strangers, ennemies, relatives, acquaintances, friends, etc. But they all share something. “Participation,” in this case, can be relatively passive and the difference between key participants (say, volunteers and lecturers in a conference) and attendees is relatively moot, at a certain level of analysis. The key, here, is the set of connections between people at the event.

These connections are a very powerful component of social networks. We typically meet people through “events,” albeit informal ones. Some events are explicitly meant to connect people who have something in common. In some circles, “networking” refers to something like this. The temporal dimension of social connections is an important one. By analogy to philosophy of language, the “first meeting” (and the set of “first impressions”) constitute the “baptism” of the personal (or social) connection. In social media especially, the nature of social connections tends to be monovalent enough that this “baptism event” gains special significance.

The online construction of social networks relies on a finite number of dimensions, including personal characteristics described in a profile, indirect connections (FOAF), shared interests, textual content, geographical location, and participation in certain activities. Depending on a variety of personal factors, people may be quite inclusive or rather exclusive, based on those dimensions. “I follow back everyone who lives in Austin” or “Only people I have met in person can belong to my inner circle.” The sophistication with which online personal connections are negotiated, along such dimensions, is a thing of beauty. In view of this sophistication, tools used in social media seem relatively crude and underdeveloped.

Going back to the (un)conference concept, the usefulness of having access to a list of all participants in a given event seems quite obvious. In an open event like BarCamp, it could greatly facilitate the event’s logistics. In a closed event with paid access, it could be linked to registration (despite geek resistance, closed events serve a purpose; one could even imagine events where attendance is free but the microblogging backchannel incurs a cost). In some events, everybody would be visible to everybody else. In others, there could be a sort of ACL for diverse types of participants. In some cases, people could be allowed to “lurk” without being seen while in others radically transparency could be enforced. For public events with all participants visible, lists of participants could be archived and used for several purposes (such as assessing which sessions in a conference are more popular or “tracking” event regulars).

One reason I keep thinking about event-specific microblogging is that I occasionally use microblogging like others use business cards. In a geek crowd, I may ask for someone’s Twitter username in order to establish a connection with that person. Typically, I will start following that person on Twitter and find opportunities to communicate with that person later on. Given the possibility for one-way relationships, it establishes a social connection without requiring personal involvement. In fact, that person may easily ignore me without the danger of a face threat.

If there were event-specific instances from microblogging platforms, we could manage connections and profiles in a more sophisticated way. For instance, someone could use a barebones profile for contacts made during an impersonal event and a full-fledged profile for contacts made during a more “intimate” event. After noticing a friend using an event-specific business card with an event-specific email address, I got to think that this event microblogging idea might serve as a way to fill a social need.


More than most of my other blogposts, I expect comments on this one. Objections are obviously welcomed, especially if they’re made thoughtfully (like my PodMtl friend made them). Suggestions would be especially useful. Or even questions about diverse points that I haven’t addressed (several of which I can already think about).



What do you think of this idea of event-based microblogging? Would you use a microblogging instance linked to an event, say at an unconference? Can you think of fun features an event-based microblogging instance could have? If you think about similar ideas you’ve seen proposed online, care to share some links?


Thanks in advance!

Intervention médiatique helvético-québécoise

Un peu la suite (tardive) d’un billet sur la «vitalité culturelle du Québec» (qui était lui-même une suite d’un billet sur le contenu québécois), avec des liens à deux baladodiffusions: David Patry (du syndicat du Journal de Montréal) en entrevue sur Musironie et Jean-François Rioux (directeur radio à RadCan) en entrevue sur Médialogues.

Un peu plus de contexte que vous n’en désirez… 🙂

J’écoute de nombreuses baladodiffusions, en français et en anglais. En tant qu’ethnographe et en tant que  bavard invétéré, j’essaie  d’apporter mon grain de sel dans diverses conversations. Certaines baladodiffusions (entre autres celles qui proviennent du contexte radiophonique traditionnel, comme Médialogues) «donnent la parole aux auditeurs» en sollicitant des messages téléphoniques ou par courriel. Une participation beaucoup moins directe ou égalitaire que dans le média social, mais une participation sociale tout de même.

En tant que Québécois d’origine suisse, je me plais à écouter des baladodiffusions helvétiques (provenant surtout de la radio publique en Suisse-Romande, la baladodiffusion indépendante étant encore plus rare en Suisse qu’au Québec). Ça m’aide à conserver un contact avec la Suisse, ne serait-ce que par l’accent des participants. Et ça me fait parfois réfléchir aux différences entre la Suisse et le Québec (ou, par extension, aux différences entre Amérique du Nord et Europe).

J’écoute des baladodiffusions de Couleur3 et de «La première» (deux stations radiophoniques de la SRG SSR idée suisse) depuis 2005. Mais ce n’est qu’en écoutant un épisode de la baladodiffusion de Vous êtes ici de Radio-Canada, l’été dernier que j’ai appris l’existence de Médialogues, une émission de La première au sujet des médias. 

Puisque je suis en réaction contre le journalisme depuis 25 ans, la critique des médias me fascine. Médialogues n’est pas, en tant que telle, représentative de l’analyse critique des médias (elle est animée par des journalistes et les journalistes peinent à utiliser un point de vue critique sur le journalisme). Mais plusieurs interventions au cours de l’émission sont effectuées par des gens (y compris d’anciens journalistes comme Christophe Hans) dotés du recul nécessaire pour comprendre le journalisme dans son ensemble et certains journalistes qui participent à l’émission énoncent à l’occasion des idées qui peuvent être utiles à l’analyse critique du journalisme.

Soit dit en passant, au sujet du respect… Je respecte qui que ce soit, y compris ceux avec qui je suis en désaccord profond. Je peux parfois sembler irrespectueux à l’égard des journalistes mais ce n’est pas contre eux que «j’en ai». Je suis en réaction contre le journalisme mais j’apprécie les journalistes en tant que personnes. Par ailleurs, je considère que beaucoup de journalistes sont eux-mêmes irrespectueux à l’égard des non-journalistes et leur manque de respect à notre égard provoque parfois en moi certaines réactions qui peuvent ressembler à des «attaques» plus personnalisées. Mon intention est toute autre, bien évidemment, mais je prends la responsabilité pour toute méprise à ce sujet. J’ai d’ailleurs été confronté à ce genre de situation, il y a quelques mois.

Revenons donc à Jean-François Rioux, en entrevue avec les journalistes de Médialogues.

Le contexte immédiat de cette entrevue est relativement simple à comprendre: la Société Suisse Romande (portion francophone de la SRG SSR idée suisse) procède en ce moment à la fusion de ses services télévisuels, radiophoniques et Internet. C’est donc un sujet qui anime et passionne l’équipe de Médialogues (située au cœur de cette transformation). La semaine dernière, intriguée par des propos de Gérard Delaloye, (dont les interventions ont été entendues à plusieurs reprises pendant la semaine), l’équipe de Médialogues s’est penchée sur la crainte toute journalistique de la perte de diversité causée par cette fusion de diverses sections du service public. N’étant pas en mesure de contacter le directeur télévision et radio (déjà sollicité à plusieurs reprises par Médialogues, à ce que j’ai pu comprendre), l’équipe de journalistes a décidé de contacter Jean-François Rioux. Choix très logique puisque la SRC est l’équivalent très direct de la SSR (y compris la distinction linguistique) et que CBC/SRC a déjà procédé à cette fusion des médias.

Rioux était donc invité à se prononcer au sujet des effets de la fusion des moyens de communication. Là où tout prend son sens, c’est que l’équipe de Médialogues utilise le terme «convergence» pour parler de cette fusion. Ce terme est tout à fait approprié puisqu’il s’agit d’un exemple de ce qu’on appelle «la convergence numérique». Mais, en contexte canadien (et, qui plus est, québécois), le terme «convergence» est fortement connoté puisqu’il a surtout été utilisé pour désigner ce qu’on appelle «la convergence des médias»: une portion de la concentration des médias qui traite plus spécifiquement de l’existence de plusieurs organes médiatiques «multi-plateforme» au sein d’une même organisation médiatique. Contrairement à ce que certains pourraient croire (et que je me tue à dire, en tant qu’ethnolinguiste), c’est pas le terme lui-même, qui pose problème. C’est l’utilisation du terme en contexte. En parlant au directeur radio de RadCan, il est bon de connaître le contexte médiatique québécois, y compris une aversion pour la convergence des médias.

En tant qu’ethnolinguiste helvético-québécois, il était de mon devoir d’indiquer à l’équipe de Médialogues qu’une partie de cette entrevue avec Rioux était tributaire d’une acception proprement québécoise du concept de «convergence». J’ai donc envoyé un courriel à cette époque, n’étant alors pas en mesure de laisser un message sur leur boîte vocale (j’étais dans un lobby d’hôtel en préparation à une visite ethnographique).

Alors que je suis chez un ami à Québec (pour d’autres visites ethnographiques), je reçois un courriel d’Alain Maillard (un des journalistes de Médialogues) s’enquérant de mes dispositions face à une entrevue téléphonique au cours des prochains jours. Je lui ai rapidement indiqué mes disponibilités et, ce matin, je reçois un autre courriel de sa part me demandant si je serais disponible dans la prochaine heure. Le moment était tout à fait opportun et nous avons pu procéder à une petite entrevue téléphonique, de 9:58 à 10:18 (heure normale de l’est).

Malheureusement, j’ai pas eu la présence d’esprit de procéder à l’enregistrement de cette entrevue. Sur Skype, ç’aurait été plus facile à faire. Compte tenu de mon opinion sur le journalisme, évidemment, mais aussi de ma passion pour le son, j’accorde une certaine importance à l’enregistrement de ce type d’entrevue.


Donc, Maillard et moi avons pu parler pendant une vingtaine de minutes. L’entrevue était proprement structurée (on parle quand même de la Suisse et, qui plus est, d’un journaliste et auteur œuvrant en Suisse). Une section portait directement sur la notion de convergence. Selon Maillard, celle-ci pourrait faire l’objet d’une diffusion de deux minutes au début de l’émission de vendredi. La seconde section portait sur mon blogue principal et se concentrait sur l’importance de bloguer dans un contexte plutôt carriériste. La troisième section portant sur un de mes «chevaux de bataille»: la musique et les modèles d’affaires désuets qui la touchent. Comme beaucoup d’autres, Maillard s’interrogeait sur les montants d’argent associés à certains produits de la musique: les enregistrement et les spectacles. Pour Maillard, comme pour beaucoup de non-musiciens (y compris les patrons de l’industrie du disque), il semble que ce soit l’accès à la musique qui se doit d’être payant. Malgré les changements importants survenus dans cette sphère d’activité para-musicale depuis la fin du siècle dernier, plusieurs semblent encore croire que La Musique est équivalente aux produits de consommations (“commodities”) qui lui sont associés. La logique utilisée semble être la suivante: si les gens peuvent «télécharger de la musique» gratuitement, comment «la musique» peut-elle survivre?  Pourtant, ce n’est pas «de la musique» qui est téléchargée, ce sont des fichiers (généralement en format MP3) qui proviennent de l’enregistrement de certaines performances musicales.

L’analogie avec des fichiers JPEG est un peu facile (et partiellement inadéquate, puisqu’elle force une notion technique sur la question) mais elle semble somme toute assez utile. Un fichier JPEG provenant d’une œuvre d’art pictural (disons, une reproduction photographique d’une peinture) n’est pas cette œuvre. Elle en est la «trace», soit. On peut même procéder à une analyse sémiotique détaillée du lien entre ce fichier et cette œuvre. Mais il est facile de comprendre que le fichier JPEG n’est pas directement équivalent à cette œuvre, que l’utilisation du fichier JPEG est distincte de (quoiqu’indirectement liée à) la démarche esthétique liée à une œuvre d’art.

On pourrait appliquer la même logique à une captation vidéo d’une performance de danse ou de théâtre.

J’ai beaucoup de choses à dire à ce sujet, ce qui est assez «dangereux». D’ailleurs, je parle peu de ces questions ici, sur mon blogue principal, parce que c’était surtout mon cheval de bataille sur le blogue que j’ai créé pour Critical World, il y a quelques temps.

Comme vous vous en êtes sûrement rendu compte, chères lectrices et chers lecteurs, je suis parti d’un sujet somme toute banal (une courte entrevue pour une émission de radio) et je suis parti dans tous les sens. C’est d’ailleurs quelque-chose que j’aime bien faire sur mon blogue, même si c’est mal considéré (surtout par les Anglophones). C’est plutôt un flot d’idées qu’un billet sur un sujet précis. Se trouvent ici plusieurs idées en germe que je souhaite aborder de nouveau à une date ultérieure. Par exemple, je pensais dernièrement à écrire un billet spécifiquement au sujet de Médialogues, avec quelques commentaires sur la transformation des médias (la crise du journalisme, par exemple). Mais je crois que c’est plus efficace pour moi de faire ce petit brouillon.

D’ailleurs, ça m’aide à effectuer mon «retour de terrain» après mes premières visites ethnographiques effectuées pour l’entreprise privée.