User:9cfilorux/Merits of IP editing

I think we all start out assuming good faith. However, when a group or category repeatedly behaves badly, or is subject to unreasonable standards of behaviour that they cannot follow, their right to have good faith assumed is collectively lost. That is how prejudices are created.

This, furthermore, is the case of IP editors: they so frequently vandalise or act in bad faith that, for many of us users, their right to have good faith assumed has been revoked. But here lies the problem with this fact of human logic. Collectivities, however unified, divided or merely categorised, do not act in any kind of faith, good or bad. Rather, it is the individuals who make them up. It follows from here that, however appealing and convenient it may seem to judge on a group level and make sweeping generalisations, we must instead judge on a case-by-case basis. Categorising, for example, all IP editors as vandals or all teenagers as messy and whiny, is rather like saying that all grains of sand are the same colour. To determine the colour of a grain of sand, one must look only at it, not at the beach; to determine what the motivation of an IP editor is, one must look only at their edits.

But still we cannot escape categorisation.

Shared IPs
Unfortunately many IPs are shared or dynamic, and there is no such 'IP editor' to speak of. In these cases, the only solution to vandalism is a soft block, the purpose of which is to separate the vandals from the serious editors. However, one can only imagine how many serious editors not interested in an account and/or without access to another computer would be blocked by a soft block. Therefore, it does not separate out only the vandals. Those whom it separates out are those who are unwilling to make an account – which coincides only imperfectly with being a vandal.

This would also be the case if we decided, as we probably will not, to block all IP edits. It is the very reason, or one of them, why IPs can still edit. We would be blocking good-faith editors, and one good-faith editor is worth a thousand vandals.

Or is it really, when we cannot tell which is the good-faith editor?

The main reason I see for disallowing IP editing is that we would not have to make so many generalisations. If everyone is forced to make an account, in theory we will be able to judge what kind of faith individual editors are editing in. This is essential in determining which edits should be reverted, and in extreme cases, prevented by means of a block.

Shared accounts
Again, however, theory and reality do not match. We may have a policy against shared accounts, but that does not stop some from slipping through.

Is this enough of a problem that disallowing bare IPs is unjustified? Maybe. Do we know? Do we know if it is any easier to tell if an account is shared than an IP?

Even if we do not, one thing is certain: the users of a shared account are far more likely than those of a shared IP to all act with the same intentions. Shared IP users most likely did not ask to have a shared IP; indeed, they may not even know about it. Shared logged-in users, on the other hand, will know of the existence of the group of people using the account; as there most assuredly is a group, one that is collaborating through the account, therefore all acting with roughly the same intentions. Therefore blocking a shared account will have a very different effect from blocking a shared IP: shared accounts are acting in concert and hence can be treated almost the same as non-shared ones. The difference is that shared accounts are more likely to be unquestionably blockworthy, not least of all because they violate a Wikipedia policy merely by being shared. One wonders why, then, it is even allowed to edit from a shared IP. Should that, too, not be against policy? It is different, yes, but the premise is the same: multiple editors editing under the same handle, which would logically include a shared IP, are being dishonest by, knowingly or not, appearing to be the same person at first glance. A possible reason for prohibiting such dishonesty has been explored above, namely that dealing with needlessly disruptive edits is made far easier when each editor has one username (or IP, but that is not feasible) and each username has one editor, because you can simply block the offender and be done with it.

Sockpuppets
Sometimes, however, the offender must be blocked multiple times – under different names. In these cases, when sockpuppetry is suspected, checkuser is carried out on the suspect socker. This situation requires obtaining knowledge of what IPs have been used so that one can block them and prevent further socks from being created.

There are two main problems with this practice.
 * When checkuser is carried out, we are judging by IP address. We do not know for sure if the truth can be found with checkuser. But we can get a decent enough idea, and if editors from that IP have never been constructive, there is no question that a block is in order. It is only if they have, a possible scenario, when we are stuck with no good solution. If good editors and bad sockpuppeteers edit from the same IP, it will be shared. In these cases account creation is disabled. The risks of disabling account creation have already been covered: a good editor may not be able to find another IP and also not have created an account by the time the block has been made.
 * Sockpuppets are usually not prevented, only stemmed. In other words, checkuser is only carried out when it is suspected that a string of socks has already been created. Does it have to be this way? We could more easily prevent sockpuppetry and the resulting needless damage if it were allowed to run checkuser on all vandalism-only accounts. This inability to simply see the vandal's IP has been cited as a reason for allowing IP editing; in fact, it may instead be a reason for loosening the rules on checkuser.

Conclusion
Whether to allow IP editing is a dilemma. Whatever we do about vandalism, we can neither control it satisfactorily nor allow all good-faith editors to edit. There are two options: allow IP edits, or disallow them; and neither is good enough.

The question, then, is which should we choose? Which has benefits that outweigh the drawbacks? We will not know until we try testing out, maybe for a year, Wikipedia without anonymous editing. This might show us whether it is a good idea. Or it might not. We'll never know until we try, will we? And we'll probably never try...