Thursday, February 18, 2010

Critique of Friedman & Currall on E-mail

Article Reviewed

Friedman, R. A., & Currall, S. C. (2003). Conflict escalation: Dispute exacerbating elements of e-mail communication. Human Relations, 56(11), 1325-1347.

Summary of the Article

Friedman and Currall (2003, p. 1327) discuss situations in which exchanges of e-mail messages exacerbate conflict.  The topic arises because of their experiences and observations of instances in which a disagreement that seemed minor and manageable escalated, via e-mail, into serious conflict, to the point of ending relationships.  They ask:  “[I]f e-mail communications that exhibited conflict escalation had occurred by phone or in person, would they have ended as they did?”  This question is important because, they say, “as the number of workplace relationships managed by e-mail increases, the implications of e-mail escalation will grow exponentially.”

The authors explain this escalation in terms of their “dispute-exacerbating model of e-mail” (DEME) (p. 1325).  Their explanation is apparently a “model” in the sense that it is “a description or analogy used to help visualize something (as an atom) that cannot be directly observed” (Merriam-Webster Online, 2010).  In this case, the model seems to consist simply of elaboration upon the view that e-mail can exacerbate disputes.  The discussion here will thus tend to refer to this as simply a view or an idea, rather than as a formal model.

Friedman and Currall identify four dispute-exacerbating characteristics of e-mail:  it provides diminished feedback, it provides diminished social cues, it can be lengthy, and it can attract excess attention from the parties.  The first two of these four seem more or less identical:  as the authors explain, e-mail does not provide an opportunity for the kinds of clues that people tend to give each other when they are conversing face-to-face or by phone.  Because of this dearth of feedback, people are more likely to misconstrue one another’s views and perceptions, especially in the direction of perceiving that the other party is behaving aggressively.  E-mail does not give people the same opportunities to understand and get to know one another, and to emphasize the importance of openness and cooperation in their interactions.  Other people are depersonalized and more easily seen as oppositional.  E-mail also provides less information about social status, along with “weak cooperative norms and weak restraint against using aggressive tactics” (p. 1336).

The authors’ third identified dispute-exacerbating characteristic of e-mail is that the ability to go on at length, in an e-mail message, carries a risk of provoking unnecessary misunderstanding and disagreement.  Long messages can be seen as rude and overwhelming.  They may result in responses that address only a few points, which may then be seen as dismissive.  Moreover, the authors say, attention may go especially to the most negative arguments, and compounded misunderstandings in an exchange of e-mails make it “harder to unravel the differences between the parties” (p. 1338).  The fourth dispute-exacerbating characteristic, excess attention, again seems to make a similar point:  people can be drawn too far into an e-mail exchange, resulting in the production of long replies and/or excessive rumination on what has been said:  “Thus, having the opportunity to focus a great deal of time on a received message may not be productive” (p. 1339).  Likewise, the opportunity to spend more time on the first message in an exchange means that one is more likely to become convinced of its truth and less likely to compromise; and the recipient is said to be more likely to assume that the sender has indeed made use of the opportunity to devote time to the message, so that its misstatements tend to be received as final statements of the sender’s firm intentions, as distinct from mere slips of the tongue.

In their conclusory section, the authors offer some advice for managing disputes.  First, they say, this is better done by phone or face-to-face than by e-mail.  If conflict management is done by e-mail, they recommend awareness to the foregoing conflict-aggravating characteristics of e-mail.  In the case of lengthy e-mails, for example, they suggest breaking the discussion down into multiple short messages, “generating as much interaction back and forth as possible” (p. 1342).  In their recommendations, they include considerations that were not, but perhaps should have been, added to the four dispute-exacerbating aspects mentioned above, especially focusing on the relationship rather than the message and avoiding “hyper-rationality” by remembering that “differences occur, and are resolved, through appropriate emotion not solely logical argument” (p. 1342).

Finally, the authors provide some caveats.  They say (p. 1344),

There are probably additional technologies that we have not covered (e.g., chat rooms), other effects of e-mail that we have not considered (e.g., the long-term impact of having written records), and other conditions that we have not explored (e.g., number of parties involved in the conflict).
They also acknowledge that e-mail is “extremely useful” and “does not turn all communications into escalated conflicts” (p. 1342), and “perhaps – over time – most people may become more skilled in e-mail” (p. 1343). 


While the authors make a number of valid points, I find that they neglect some important characteristics of e-mail in conflictual situations.  The following paragraph address those characteristics briefly.

First, a bit about my perspective.  I started using RSTS e-mail at Columbia in the late 1970s, and continued with CompuServe and other e-mail services in the 1980s and thereafter.  In the early 1990s, I worked at a federal agency that utilized e-mail to connect people in a half-dozen regional offices around the country.  Later that decade, I was active on Usenet newsgroups via Deja News and otherwise.  This background means that I have had many opportunities to make the sorts of mistakes that Friedman and Currall describe, and also to be on the receiving end when others have done likewise.  So when I review this article, I have reactions on both sides of the matter.

A basic assumption in this article is that the conflict, whatever it is, should be resolved rather than exacerbated.  Resolution, for Friedman and Currall, seems to mean softening and downplaying, with the aid of social norms and cues, so that a person will not say something via e-mail that s/he would not have said in person.  This makes sense in the context of an example like the one with which they begin the article, where a conflict unnecessarily spiraled out of control to the point of terminating a relationship.  But it also raises a number of questions.

One question is why the relationship in that instance was terminated.  The relationship in question apparently involved an editor of a scholarly journal and a writer who had submitted an article.  The conflict seems to have involved revisions requested by the editor.  According to the authors (p. 1326), “Each side had been presenting arguments back and forth until the editor . . . e-mailed that he was ‘ending our relationship.’”  In that instance, e-mail seems to have been facilitated the exchange of views between people who disagreed.  Should the parties have attempted to conduct that exchange via telephone?  Friedman and Currall are probably correct in implying that it would not have continued for long via phone – but why not?  The answer is, surely, that the editor occupied a position of power vis-à-vis the writer, and in a spoken conversation the writer would have received cues that the editor was displeased.  Being a sensible person, the writer would have had to take a submissive posture and allow the editor to have his/her way.  Instead, what may have happened was that the writer’s opportunity to think about and express his/her arguments via e-mail produced some conviction that those arguments had merit, and the editor was simply not willing to acknowledge his/her error.

This example provides a dubious guide for social workers.  First, it is not at all clear that social workers should invariably take a submissive posture when they confront people in positions of power about matters in disagreement.  Moreover, the opportunity to articulate and present one’s concerns can be enormously empowering, especially when one would be afraid to do so in conversation, or would simply lack any opportunity to do so because of cultural or organizational barriers.  For example, an e-mail may get forwarded to a top executive in instances when the writer him/herself would not be invited to speak to that executive.

Friedman and Currall (p. 1331) say that e-mail will escalate the level of conflict if it encourages the use of more aggressive tactics, produces a change in one’s view of the other, weakens interpersonal bonds, or makes problems more difficult to resolve.  This is precisely what e-mail does, they say, via the dispute-exacerbating characteristics cited above:  diminished feedback and social cues, and excessive verbosity and attention to detail.  But there is a straw man here.  These remarks assume a quality of face-to-face interaction that is often nonexistent.  There are people who dislike one another, who are bullied, or who for other reasons find it unpleasant to interact with another person face-to-face.  There are jealousies and unintended as well as intended offenses in remarks that people make to one another.  It can be harder to save face and maintain self-respect when an aggressor chooses to deliver such remarks in a context where the recipient has no opportunity to respond.  Again, contrary to what these authors say, interruption and denying the other person a chance to explain him/herself is more likely in face-to-face interaction than in e-mail, where one generally has at least the possibility of expressing oneself.

What seems to happen is not that e-mail somehow unmasks an evil side within each of us.  The more likely scenario is that, when there is a disagreement, it is attributable – as Friedman and Currall say – to a departure from a social norm.  But social norms are not intrinsically good.  They can easily incorporate or support aggression, prejudice, violence, and other types of dysfunctionality.  Much of what the authors say about e-mail is compatible with the possibility that e-mail permits people to present their concerns in ways that some such norms suppress.

One important assumption in the argument presented by Friedman and Currall is that social norms are restricted to face-to-face interactions.  But there has not been, in fact, the nationwide escalation of e-mail-based conflict the authors seem to fear.  People have become accustomed to e-mail and have developed new norms around it.  Sitting at a keyboard does not liberate one from social norms; more likely, it just invokes an adapted set of norms.  As in other kinds of interactions, these norms are not homogeneous:  some people will say things that others would not, each for reasons of self-protection, articulation, comity, or whatever they consider important in the particular situation.  These choices seem to reflect what people have found feasible and appropriate in their use of e-mail.

There is also a collective aspect to conflict, insofar as the straightforward presentation of concerns by one person may strain relations between the sender and the recipient, but may also nudge the recipient toward being more receptive when others present similar concerns.  In this case, to put the shoe on the other foot, perhaps the writer finds it difficult to learn from editorial criticism.  If so, the experience of running full-tilt into several brick walls may provoke some salutary reflection and growth, where the experience of simply giving in to the person with superior power may leave him/her muttering under his/her breath, ready to make the same mistakes in his/her next editorial interaction.  Conflict can be instructive.  Even in work contexts, and even when it leads to ruptures, it is not always a bad thing.

There is also an efficiency concern.  Editors frequently disagree with writers on various points.  The anecdote would have been more telling if it had been provided along with information about the larger context.  If the editor handles a hundred authors within a certain timeframe, perhaps a small percentage of those will carry disputes to an extreme, but most presumably do not, else editors would probably not use e-mail.  In the anecdotal situation, the editor probably did not need that particular author very badly, given that journal editors tend to be buried in article submissions.  What the writer apparently did not keep in mind, as s/he was crafting his/her replies, was that s/he was dispensable in this particular arrangement.

These observations highlight the educational aspect of e-mail conflict.  Turnage (2008) lists characteristics of e-mail messages that are highly likely to be considered “flaming.”  These characteristics include angry uses of profanity, all capital letters, and excessive exclamation points or question marks.  As suggested by the foregoing remarks about social norms, people tend to learn that these sorts of messages are ineffective if not counterproductive.  As users have become more familiar with flaming, they seem to have become more individually and collectively resistant to it.  In online discussion forums, for example, disputes featuring such behaviors tend to be abandoned by people who have no strong interest in either side of the argument, leaving those who do, along with those who would mediate – which may be an entirely appropriate outcome, especially if it does facilitate interaction among people of opposing who would otherwise be preaching to their own respective choirs.

I’ll rest that line of argument with the general statement that e-mail conflict management is more nuanced and valuable than the authors allow.  On a different level, it seems important to report the results of a bit of fact-checking.  I looked into a 1991 article by Clark and Brennan entitled “Grounding in Communication.”  Friedman and Currall cite this article for the view that grounding is the process in which parties to an interaction develop shared senses of understanding and participation, and “this is important because ‘speech is evanescent’” (p. 1328, quoting Clark & Brennan, p. 128).  This, it turns out, is not an accurate report of the earlier article.  Contra Friedman and Currall, the 1991 article states, not that there are “six tools for grounding” (F&C, p. 1328), but, rather, that there are eight constraints that a medium may impose upon people (C&B, p. 140).

Friedman and Currall do acknowledge that e-mail has an advantage in the last two of those eight, reviewability and revisability, but they don't make clear that Clark and Brennan differentiated “grounding in conversation” (p. 128) from grounding in other sorts of communications.  Clark and Brennan cite the “keyboard teleconferencing” of their time as an example where grounding was indeed attainable, although acknowledgments were more “costly” because they would take more effort (because, in their era, “it is difficult to time an acknowledgment precisely, and trying to do so may interrupt the other typist”) (p. 140).  That is still a problem in chat, but much less so, because chat is very fast now, and its users tend to become used to it.  At any rate, acknowledgment is probably less of a problem in e-mail than in face-to-face conversation, where people often do not show that they have even heard, much less comprehended, what the other party has said.

On the fact-checking level, it also seemed to me that Friedman and Currall somewhat distort the matter when they criticize e-mail for permitting the use of certain “tactics” that are not available in ordinary face-to-face communications.  One is what they call “argument bundling,” where “e-mail comments can be very long and include multiple points . . . . without the receiver having the opportunity to respond or clarify” (p. 1329).  Of course, speakers at conferences and rallies have been doing this for millenia.  It is not a tactic; like e-mail, it is simply a mode of communication, with its own significant advantages for some purposes.  The difference is not its one-sidedness; it is that recipients tend to be self-selected to agree in the public meeting context.  That difference is important and interesting, but it is not explored in this article.

Actually, recipients of long e-mail messages do have opportunities to respond in ways of their preference.  For instance, they can break long messages into smaller pieces and explain that, for lack of time or interest, they are responding, but for now, as a courtesy, will respond at least to the first several sentences or paragraph of the message they received.

One final fact-oriented rejoinder:  e-mail, according to Friedman and Currall, is “profoundly asocial” (p. 1329).  Were this true, social media would have been stillborn.  The more accurate statement, I suggest, is that e-mail is social in ways that are unfamiliar to those who have not yet added it to their repertoire.  One thing the authors consider particularly asocial about e-mail is that “E-mails are typically received and written while the writer is in isolation, staring at a computer screen – perhaps for hours at a time, so that awareness of the humanness of the counterpart may be diminished” (p. 1329).  This is not necessarily true; e-mails are typically written in a variety of situations and, when they involve the workplace setting that the article sometimes cites as its focus, isolation is atypical.  Staring at a computer screen may or may not have the effect of diminishing awareness of the humanness of others; but if it does, this problem goes well beyond e-mail to the entire arrangement and even the viability of many contemporary, computer-dependent workplaces.  Indeed, if awareness of humanness is to be prioritized, it is not clear that the coldly efficient market economy itself can survive.

The article by Friedman and Currall turns out to have been interesting but unidimensional.  Email and other electronic communications do have drawbacks in conflict management situations.  There will be conflicts that go bad in e-mail, that would have turned out better in other media.  This is true of all forms and styles of communication and interaction.  People are best advised, not to run pell-mell away from e-mail when conflict emerges, but rather to become more familiar with its advantages and to implement strategies that help them to minimize its drawbacks.

Wednesday, February 10, 2010

Ubuntu 9.10, VMware Workstation 7: Failed to Open Sound Device (Workaround)

In previous posts, I tried fixing error messages that I was getting in a Windows XP virtual machine (VM) running on VMware Workstation 6.5.2 on Ubuntu Linux, version 9.04.  I didn't know, at that time, whether some of my problems stemmed from having done an upgrade rather than a clean install of Ubuntu.  Now I found myself facing the same problem again, after a clean install of both Ubuntu 9.10 (Jaunty Jackalope) and VMware Workstation 7.

The problem was as follows:  if I checked the "Connect at power on" option, then audio would not run in the VM until (in full screen mode) I went into VM > Removable Devices > Sound Card > Connect; but when I did that, I would get this error message:

Failed to open sound device /dev/audio: Device or resource busy.  Failed to connect virtual device sound.

One recent discussion seemed to suggest that the problem was that Flash Player and/or Alsa audio.  I suspended my VMs and rebooted the computer.  When it rebooted, I made sure that Firefox was not running.  I resumed the VM and audio played OK in IrfanView.  I started Firefox inside the VM (i.e., in Windows XP) and played a YouTube video.  The audio was still OK in IrfanView.  I started a second session of Workstation and resumed a different VM in that session.  The manual sound card connection went OK; and there, too, I could play audio without stuttering.  The audio in the second VM was not as good; it had some static.

With those two VMs open, I started Firefox in Ubuntu (i.e., not in a WinXP VM).  I played another YouTube video.  Now I got an error message, and audio would not play in IrfanView within the VM.  I tried the other VM; same thing.  The error message was "Failed to open sound device /dev/audio: Device or resource busy.  Sound will be disconnected."  I went back to Ubuntu and killed Firefox.  Now I was able to connect the sound card and play audio inside the two VMs, same as before, complete with mini-stutters in the second one.  I started Firefox again in Ubuntu.  The audio was still OK inside the VM.  If I was playing audio in IrfanView in the VM and then went into a YouTube page in Firefox, the audio would not play in the latter.  I had to restart Firefox to get its YouTube audio to play, and then, as before, I was not able to hear audio inside the VM.  It worked the same way if I played a YouTube video in Opera rather than in Firefox.  So it seemed that the problem (which has apparently been around for years) continued to be within VMware Workstation.

For the time being, the solution seems to be either (a) to watch videos and other webpages that use Flash, do it in a browser session that is running inside your VM, not in a browser running in Ubuntu, or (b) after watching a video or otherwise using Flash in Ubuntu, kill the program that used it (e.g., Firefox) and manually reconnect with your sound card inside the VM.

Saturday, February 6, 2010

Error While Resizing Vista NTFS Partition with GParted

I was using the GParted 0.5.1 live CD to resize an NTFS partition.  (I had already backed up the data on the partition.)  When I tried to resize in GParted, I got an error:

An error occurred while applying the operations

See the details for more information.

if you want support, you need to provide the saved details!
save_details.htm for more information

I clicked OK.  Unfortunately, the details did not explain what had gone wrong.  I rebooted into Vista and then did a complete shutdown.  (Previously, I had just hibernated the system.)  I allowed a minute or so for the memory to clear, and then rebooted with the GParted live CD.  This time, it worked.  The problem was that I had hibernated instead of completely shutting down Vista before using GParted.