Sexual orientation is a logically inappropriate category: each of its 3 values is a doubly misleading improper concept conflating orthogonal categories and assuming unwarranted parallelism. From this arise homophobia and homosexualism, both irrational. I have been bawled out for complaining about having to code for this inappropriate category, but we must be brave in the face of bigotry. I do not know what if anything to put in place of this improper category.
I think that sexual orientation should be kept - I don’t see any compelling argument here.
OK, I’m going to be bold and rewrite this section. I’d like it to be more positive - tell you what to do as much as what not to do. How about this:
1.4 Strive for inclusiveness and accessibility in every aspect of the technologies you create.
Technologies should be designed to treat all people with fairness and respect. Inequities between different groups of people may result from careless or intentional misuse of information and technology. To combat this, accessibility and social consequences of new technologies should be considered at every phase of the product life cycle, from design through development, to eventual product retirement.
Prejudicial discrimination on the basis of superficial attributes, stereotypes, or similarly inappropriate factors is a violation of ACM policy. Harassment in general and sexual harassment in particular is not appropriate behavior for ACM members and will not be tolerated.
Update 2017-06-15: I removed the laundry list of common types of prejudicial discrimination and replaced it with more generic language.
What are the chances that that audience doesn’t contain a blind, hard of hearing, or disabled person? Fundamentally, accessibility is a critical part of non-discrimination. In fact, that part should probably be written out here a bit (this seems to be the best place to put a statement on accessibility.)
I think that “gender identity” needs to be deleted from 1.4 for the following reasons:
Many consider that men believing that they are women or vice versa (in conflict with XX vs. XY genetic evidence or well-accepted physical attributes) is a mental illness just as someone believing that he or she is Napoleon or a dog is mental illness.
One person’s rights end where it conflicts with someone else’s rights. As a man I felt violated when women have entered the Gentlemen’s Room. I expect that women would suffer more severe violation and fear of rape if a man entered the Ladie’s Room, locker room, etc. The potential for abuse by men entering the Ladie’s Room “for a peek” seems quite high as well.
It is quite inappropriate for a professional organization or those entrusted with leading it to act as the extension of one political party or end of a political spectrum. Such an extreme and divisive issue should not even be considered without a vote of the membership.
This proposed requirement on members conflicts with the law in many U.S. states and those of other countries w.r.t. bathroom and locker room access. Asking a member to violate the law is grossly inappropriate and risks legal consequences for ACM and its officers.
I think that “religion or belief” should be changed to “religion or lack thereof”. The term “belief” is too general and can lead to abuse such as “I believe Hispanic women are superior to white men” being accepted when clearly this is prejudice. Of greater import, those who are not religious should be protected as well.
Lastly I agree with others that “, or any other inappropriate factor” should be deleted both because it is vague and thus lacks guidance and because it can be abused. One person’s inappropriate factor is another’s prejudice.
Given that English is probably not the first language of many members of the ACM, I would be hesitant with using words that are not that well-known.
I apologise that I only just saw this thread and comments close in today, so I can not properly situation my comment in terms of all of the prior arguments made.
I would like to see a mention of all ACM events being accessible to people with disabilities, and all papers be accessible. In SIGCHI we have been tracking this for our own events, and we still do not even have an accessibility chair for some of the conferences.
Often I can not attend conferences or read articles in the ACM because they are not accessible. Folks say it is not worth the time or financial costs to do so given how small a population it impacts. Yet, it is catastrophic to the careers of disabled scholars, and we need disabled CS folks to ensure better accessible tech based on first hand experiences.
This revised statement is a lovely idealised statement of values, but their is no enforcing any penalty for failure to follow it, or specification of domains it covers. Thus, this doesn’t represent an improvement over the old statement with regards to accessibility.
Dr. Jennifer A. Rode
Blind, hard of hearing, or disabled person wouldn’t be authorized to be in a number of different context think about some industrial jobs, about LE/Defense, transport. There is also the case of people starting something: are you going to spend the time implementing and testing accessibility features or are you going to work on that extra feature, fix that extra bug that could make the difference for you.
Many people have responded that “or other inappropriate factor” is vague. That is intentional. In fact, I’d like to shorten the list to, “Prejudicial discrimination on the basis of superficial attributes, stereotypes, or similarly inappropriate factors is a violation of ACM policy.”
The list of discrimination will never be complete, nor is one list suitable for every environment:
Prison software helps one group of people restrict another group’s rights. Software is used in mental rehabilitation, or for patients with dementia. Such software must discriminate between murderers and jailers, or between doctors and patients. The point is, when working on such software is to ask about every feature, “Is this feature rehabilitative, or punitive?”
Forum software for people who are recovering from hysterectomies is for women only. Excluding women from a forum for men recovering from prostrate surgery is also fair.
Being “inclusive” means something very different with a single-developer startup than it does at a company with thousands of developers.
I think our job here is NOT to make laws and punishments, or to define appropriate actions for every situation, but to set a direction. We want to foster reasonably open and fair meetings and collaboration. We want to select people because they are right for the ACM, not because they look or think like me.
Eric Raymond once said, “Given enough eyeballs, all problems are trivial” but that’s only true if the eyes are looking from different perspectives. A variety of perspectives is the best thing that can happen to an organization like the ACM. Our policies need to promote that. Our primary objective is still “Advancing computing as a Science and Profession” and diverse membership is one of the ways we achieve that.
The ACM can say, “Try to be fair. Strive to look beyond your own prejudice.” I think the more specific we get, the more trouble we’ll run into. Point a direction. Don’t fill in the details because there is no single right answer for every situation. Sometimes the best we can do is the least-wrong answer.
If there is an issue, the ACM might say, “You know, X generally works in this situation better than what you did” and start a dialog about improvement. What’s got to be off the table is where a chapter or ACM member knows that someone is really doing something really evil, and you shelter them. We need to point enough of a direction that sheltering bad behavior is clearly off limits.
I don’t understand these comments at all.
The examples you give don’t make much sense. What do you mean by “prison software” or “software used in mental rehabilitation”? Your statements about “discriminating” between murders and jailers or doctors and patients seems to be an issue of authentication and/or authorization and not discrimination. I just don’t understand where any of this is coming from.
The second example of forum software…is the software excluding people? Or are the administrators using general software and creating policies that exclude people? There’s a distinct difference. I highly suspect that someone did not custom build software for a forum that excludes people. Instead, they created a culture of exclusion (which may be right or wrong - I believe you pointed out two cases where you may rightfully want to only include a portion of the population from your social software).
I generally agree with the rest of your post (although I object to “selecting people” for membership in the ACM). I also think that there is a need for “laws and punishments” for membership in the ACM, otherwise membership loses value if you cannot have sanctions or other actions against people who do not meet a minimum standard of behavior. Both rules for a minimum standard of behavior as well as something that can be used to start a dialog are needed, but the same thing cannot cover both.
With respect, since many people have responded that it is too vague TAKE
IT OUT! ACM’s Ethics, like laws, must not be vague lest they be abused!
Regarding your example of a forum on hysterectomies blocking men, that
might be discrimination as well. As a man helping my then wife recover
from her hysterectomy I might be more justified in being on that forum
than a woman not having to deal with that.
Robert (“Bob”) Toxen
Software can be used by the police for various tasks, and (AI especially) typically absorbs any prejudices in the training dataset. Then, through no real intention of the software developers, data scientists, or anyone else involved, the software makes discriminatory decisions. This is a really common pattern, and something that, ethically, we need to be on the lookout for. (To the point where, if you are using AI for anything serious, you probably need to figure out some way to test for this.)
When it comes to rules, they cannot exist without principles. The thing that you “use to start a conversation” sits at the root of every legal system worldwide. In designing a code of laws, the first critical step is to identify general principles upon which you will build code. This is why the United States (and many other countries) have a constitution. It is also why those constitutions often contain portions that are amenable to interpretation. Additionally, these intentionally ambiguous documents typically have the force of law, and the principles that they contain are often used to detect and override bad laws.
Given world events of the past two years, I think we need to add something about “political views” or “political philosophy” to the list of areas to not discriminate against. Perhaps between " labor union membership" and “military status” might be appropriate.