[Air-l] Re: ISPs, state action, and public forums

[Bram Dov Abramson]

Sandra brought out some interesting points -- three threads, I think -- on ISPs and e-mail:

>ISPs now function as quasi-public forums because
>they are fundamental to the ways in which we
>communicate via the internet.  As quasi-public
>forums, constitutional standards must be met in
>order to restrict speech. I'm arguing that public
>forum analysis must be applied to ISPs.  This, too,
>has been raised in court but earlier on, before
>terms of service began to become harmonized with
>each other, and before the internet had become
>so significant as a medium of public communication.

1) Parenthetic, maybe, but we're talking about two separate services, right?  There's Internet access.  And then there's Internet e-mail.  As is traditional to telecom (cf PSTN), the commercial ISP model evolved around a bundled service offering -- most Internet access customers expect to get both the basic connectivity, and then the e-mail service on top.  As is not traditional to telecom, Internet connectivity is a more successful service than any of the applications built on top of it.  But that's the magic of the Internet.

(Both Internet access and Internet e-mail is itself a bundle, of course -- Internet access usually including simple IP routing and ICANN-compatible DNS, e-mail usually including at least POP and SMTP. Latour enthusiasts, insert quote here on material things being actor-networks and choosing to talk about things being a matter of choosing where the buck stops, or where the analytic line is drawn.)

I underline the distinction in this conversation because as users we want, and as researchers we think we see the benefit in, being able to mix and match those services.  I buy residential Internet access from the same company that sells me cable TV, but I don't buy my primary inbound e-mail service from that company -- although they bundle such a service with what I'm already paying for.  The value of non-ISP-dependent e-mail, or at least e-mail addressing, exceeds the extra resources it costs me to get it, in other words.  (This opens into thinking about the pros and cons of bundled e-mail having become part of the ISP model -- what sorts of e-mail services might have become standard if nobody expected to already have paid for basic e-mail whenever they paid for Internet connectivity? But that's a separate, and kind of nebulous, conversation, full of what-ifs.)

2) So let's say, as it sounds like you're suggesting, that e-mail, or maybe entire classes of Internet application ("media") including but not limited to e-mail, are "public" in the way that the PSTN is public.  This is sounding a lot like common carrier obligations.  Not necessarily, you say; there are some alternatives:

>- forming groups of users (such as all libraries,
>or all universities) large enough to negotiate preferred
>terms of service with ISPs (in the way that large
>business-oriented buildings get their own terms).
>
>- identifying protection of free speech as a market
>niche for an ISP that will choose to market itself
>that way.  (with the problem, as we've recently seen,
>that such an ISP must have its own backbone or
>there is the possibility of being shut down by the
>backbone operator).
>
>- education of users to generate a shift in public
>attitude regarding what is acceptable in ISP
>terms of service and acceptable use policies.

... all of which work.  One halfway step that has probably been bounced around -- half way, that is, between the legal remedy of common carrier obligations (but isn't that heavyhanded?), and the market remedy of just do it (but are there political-economic forces which make just-do-it not viable?) is to link this debate with the one on ISP obligations over content -- where they originate or terminate illegal content, are they liable? -- and carve out a common-carrier-like status armed with a precise definition of what it means to be such a common carrier.  ISPs who chose to adhere to this status, and not discriminate between packets, would by definition be protected from liability; ISPs who chose to discriminate between packets would, similarly, not be protected from liability.

There are a bunch of problems with that halfway step.  For Internet access, defining non-discrimination is hard, but I think that's a solvable (assuming ongoing tuning) problem.  Whether the model transfers well to the various components of e-mail service, I don't know.  And, in any case, it's not obvious that the just-do-it remedies aren't workable -- or, at least, can't be made workable in the presence of other ways of dealing with things like, say, public funding.

So, in the meantime, I very much agree on the education part, or at least on putting out there the idea that depending on one's Internet access provider for primary inbound e-mail may one day make one unhappy, and that the terms of service for one's e-mail service may one day make one unhappy, too.  On the other hand, agreeing with that means that I disagree that it's useful, right here, right now, to saddle the Internet with the legal status that the PSTN has, as a public communications service with a complicated machinery of stabilizing factors that the state puts into place.  Or does "public forum" sidestep those implications?

3) Finally:

>David Silver comments that many folks he knows are opening
>yahoo or other accounts separate from those they hold
>at universities and/or their employers and/or with smaller
>ISPs because of fear that political speech during these
>times may be restricted in those other venues.

and

>-- but to bring the conversation back to where it
>started, which was the assumption of individuals 
>who are politically active that switching to AOL or Yahoo
>in hopes of finding a more open speech environment
>-- this remains an unfounded assumption.  

I think that, for many, this is as much about e-mail addresses' signifying power in the, uh, discursive construction of organizational identity, as with the institutional power to curtail e-mail speech tout court.  Certainly, that's my case.  The reason I use this e-mail address and not that of my employer is because, as they say, my opinions are my own and do not necessarily reflect those of my employer, and nobody should be given the impression that it isn't so.

Now, in my own case this is a bit more acute than in that of, say, university-attached folks, because my employer is a communications regulator whose legal status is that of an administrative tribunal (ie a court), and giving false impressions of partiality aren't such a good idea.  To whatever degree, though, this is certainly a motivator for many: the question of *which* e-mail provider is downstream from the question of simply finding a third-party provider in the first place.

cheers
Bram