Hacker News new | past | comments | ask | show | jobs | submit login
Schneier: The Dishonest Minority: Security and its Role in Modern Society (schneier.com)
171 points by FilterJoe on May 9, 2011 | hide | past | favorite | 57 comments



"But none of these systems, with the possible exception of some fanciful science-fiction technologies, can ever bring that dishonest minority down to zero."

This is a concept that is very useful to internalize. In systems its the 'no subsystem will operate 100% correctly all of the time' in social justice scenarios its 'there will always be people who are poor'.

While the first part of that "things break" or "shit happens" to be colloquial, is accepted easily the latter 'there will always be poor people' is less easily so. But as Schneir has eloquently pointed out in his thesis statement humans are just another system composed of independent actors who are nominally out for the collective good.

Understanding this core systems concept can really help you understand where you will need to focus systemic processes or tools which will maintain the systems primary function.


I fail to see what connection the article has to poor people. Poor people are neither "uncooperative" (indeed, they tend to occupy some of the most cooperative jobs where they are told what to do and then do it), nor "parasitic" (they tend to work very hard at jobs that benefit society), nor do they deviate from the "social norm." Nor, in many societies, are they a minority.

I suspect you have confused poor people with criminals, dissenters, or activists.


No, he's just shifting from Schneier's original point out to "no system will operate 100% correctly", then from there to economic inequality.


Schneier's original point is "every cooperative system will always contain parasitic elements; while these elements can be deleterious to the correct functioning of the system, they are also a critically important source of diversity and change."

If all the reader gets out of the article is that "systems never perform at 100% efficiency" then they are ignoring the entire point of the article. As such, the root post seems to be hijacking the topic rather than talking about what Schneier actually said.


I got three messages out of Schneir's thesis:

1) In any system of nominally co-operative elements, with ability for individual elements to flip between co-operating or defecting (to use the same game theory terms Schneir did) will have some fraction of those elements in 'defecting' mode.

2) He accepts as fact that perfect control over the choice of defecting or co-operating is impossible to achieve in the systems he's talking about.

3) He outlines a way to reason about the processes and measurements one can put into a system which maximizes for 'cooperation'.

Then Schneir applies this very good and universal tenet toward security. and finally points out that even if systems could achieve perfect control that would be net negative as the 'defectors' in the system can provide the mutation/change which improves the system overall.

It is a solid piece of reasoning which, in my opinon, applies to all systems, not just security systems. Systems that include groups of people, computer networks, air traffic control, commodity markets, Etc. All of them.

Take a Gaussian distribution and point at one of the wings and there be the 'dishonest minority' there. If you are a designing a system where the actors (or agents) in your system have the freedom to choose their own path through it, whether it is an MMORPG or a botnet or a homeless shelter, you have to consider (and prepare for) the fact that a portion of your system's population will be working against the system as a whole and if you don't want the whole system to collapse you need to heed Schneir's advice and layer in processes which 'self heal' or 'detect and deter' such defections. Otherwise your system will become unstable, and then collapse.


It seems like the point the OP was driving at is that Schneier's statement is a subset of that maxim. While Schneier's point is certainly important, the generalized form of "systems are not perfect, work with it" is more widely applicable. He wasn't ignoring what Schneier said, but agreeing with and building on it.


The poor are a parasitic element, if we're talking about the same "poor". I'm not talking about the hard-working folks who bring in a crappy wage because their skillset sucks, I'm talking about the folks who have decided that it's a better idea just to stand on a street corner with their hand out hoping to take advantage of the compassion of others.


If by poor you mean lazy, then say lazy. Right now you are poor in karma because you were too lazy to articulate your idea more clearly, and have obviously offended several people enough to get downvoted. I wasn't one of them, but statements like 'the poor are a parasitic element' are pointlessly antagonistic.


While I'd clearly be wrong to state that there are no such people, I think it's important to note that even most people with the signs asking for money aren't doing so by choice. Once you're in a state where you really do need to beg for money for basic sustenance as such (which can happen surprisingly easily through no real fault of your own), it's really, really hard to get out, no matter how hard you try. And, in the meantime, while you are working to get out of it, you have to do something to live.

There is some (insufficient, at least in the US) help available from government programs and whatnot, but even non-participation in such things is not always necessarily by choice.

I've known a fair number of people who've worked their way out of poverty (my family among them), and I've known a larger number who have failed to do so by no real fault of their own. Your unkind generalisations are most definitely unwelcome.


I don't think that is an appropriate generalisation, because you are drawing a similarity between the game theoretical outcome of properties on systems of rational actors and the outcome of properties due to random physical occurrences ('cosmic rays' etc...) and unintentional irrational behaviour ('human error').

While both can result in outcomes that are collectively undesirable, I don't think useful inferences can be made by making a vague generalisation by pointing to two different classes of system (especially not without defining the generalisation unambiguously and proving rigorously that it is always true).


That's what Jesus said, "the poor you will always have with you"


It's interesting that so many commenters take issue with the term "dishonest minority." I agree that honesty is usually the best policy, but I would lie to the Nazis about the Jews in the addic. This is the value in the dishonest minority and the conflict that the term implies to people who equate dishonesty with evil is central to the point.


Schneier is Kantian, here. He probably has read the "Critique of practical reason." Let me tell you of the debate between Benjamin Constant and Immanuel Kant in 1797 IIRC, after the Critique's publication.

Kant affirmed therein that lying, as a general rule, is always wrong. Constant contradicted this idea, asking : how should you manage the case when someone comes to you and ask for protection from some bandits about to kill him, then? Shouldn't you tell the truth to the violent mob, that the guy they're looking for is indeed hidden in your closet?

Kant replies that the right thing to do is to lie to protect the innocent, but knowing that it's as a constrained exception to the absolute rule : "you should not lie". The fact that you commit a wrong for a greater good doesn't entirely excuse the wrong or alter the rule.

Most people, however, fail to adhere to this subtle analysis. Hence their displeasure with the word "dishonest".


The issue that I have with the term 'dishonest minority' is that it isn't accurate in the groups that he is attempting to describe. I love his underlying concepts, but inclusion of all rule-breakers as 'dishonest' is problematic.

The issue isn't whether lying is wrong or right. The issue is whether breaking a man-made rule is always wrong or right. "Dishonest" is not as accurate as "insubordinate" when including all murderers who break the law in with someone who blocks traffic to protest civil rights injustice. They are both breaking the rules, but is the second person actually 'dishonest' or rather, non-compliant?

The overall issue is that it's pretty clear the term 'dishonest minority' was chosen for the shock value, not accuracy. In a technical book that seems so promising, using a carnival barker tactic to raise interest seems to belittle the underlying content.


Schneier himself makes the same point even more forcefully. Quoting:

"The term 'dishonest minority' is not a moral judgment; it simply describes the minority who does not follow societal norm. Since many societal norms are in fact immoral, sometimes the dishonest minority serves as a catalyst for social change. Societies without a reservoir of people who don't follow the rules lack an important mechanism for societal evolution. Vibrant societies need a dishonest minority; if society makes its dishonest minority too small, it stifles dissent as well as common crime."


The usage of the term 'dishonest', is, by very definition, a moral judgement. Saying a word means something else doesn't make it true.


But if the author redefines a word within the context of an article or other piece of work, it seems a tad trite to continue to complain about how it's use doesn't agree with some standard definition.

So yes, I would argue saying a word means something can make it true with the context of the author's writing. It basically becomes variable declaration at that point.

Do you disagree?


Yes, I disagree. Because what you are describing is what's known as 'informal logic'.

If I can convince someone that something is true, such as "red is equal to blue", even when it isn't, then it allows an entire logical argument to be built on that falsehood. The argument may be perfectly sound as long as everyone agrees with the original falsehood.

If I say 'red is equal to blue' and you agree, then I can follow up and say, 'bulls must, therefore, get angry when they see the color blue'. This is a silly example but it's commonly used in politics and newspapers daily.

I might accept that an author could redefine a word temporarily, in some instances of fiction; however, never in a technical or nonfiction work could this be considered acceptable or even credible.


I guess I'm not understanding your perspective.

From my perspective it seems you are being overly rigid when you are complaining about the usage of the word 'dishonest'. And I don't find your example useful because I believe it's triviality ignores what makes arguing over the definition of a word difficult in the first place.

Plus, from my perspective, I'd have no problem with the statement "bulls must, therefore, get angry when they see the color blue" if "blue" was previously redefined as meaning "red".


Not to nitpick, but bulls are color blind...



Except its not, at least not as the term honest is strictly defined. To quote mirriam webste, honesty is being free of deception. While it is often immoral to decieve, there are definitely exceptions to this, which is precisely the point. You can be deceptive, and thus dishonest, while being highly moral, at least in cirtain circumstances.


You can't sever a word from its connotations just by pointing out exceptions.


> The overall issue is that it's pretty clear the term 'dishonest minority' was chosen for the shock value, not accuracy.

Yes, indeed. Either Schneier displays his old, astute, successful writer wit at work; or maybe it's his evil editor's influence :)


I agree that the title could use some work, notwithstanding the point about unjust laws. Maybe he needs to find a more neutral angle (eg The Disobedient Minority). Or maybe he could make his point more forcefully by playing up the moralistic aspect; The Wicked Minority would be provocative, but acknowledge both the issue of truly bad actors who present a genuine danger to their neighbors, and the equally worrisome problem of moral hazard that results from adopting a moralistic approach to a practical problem (ie the police state).

I admire Schneier and am interested to see where he goes with the book, but the thesis as presented seems a bit thin.


We're talking edge cases though. Most of us don't live in Nazi Germany, and most of us don't need to go round lying or breaking the law in order to be good.

Besides, it's clear that the "dishonesty" he's talking about here is less about "telling lies" and more about things like theft, fraud, and other ways of illegitimately enriching yourself at the expense of others. This is the form of dishonesty in which he's interested, because it's so central in human relations... much of human nature is designed to cope with this problem. This thread has got rather off-topic with too much talk about how we classify Rosa Parks, who is fairly irrelevant to the problem at hand.


Insubordinate is a much better term.


Who cares? Use fewer words or say something interesting.


Morality's purpose is to further human life, not result in the sacrifice of it.

Its quite ok to lie to a thief or to protect someone that you care about. Or even a random stranger.

These are exceptions but it is in the interest of those choosing to be honest with those who are honest with them to make such exceptions.

Giving the benefits of honesty to those who refuse to practice it is simply enabling the parasitism.


The word "dishonest" comes with a baggage of connotations. Most people aren't going to read the book and they'll misunderstand the purpose.

To give an example : as much as I think that "The Selfish Gene" is an awesome title, I feel like I have to defend the choice of words whenever I recommend the book. It doesn't really convey the meaning correctly. It means the right thing if you've read the book; otherwise it's misleading.

That's what I think will happen with the term "dishonest minority".


The problem with the Dawkins title is that there are two ways to parse it. The correct (imho) way is to read it as a declaration that genes as a class are selfish. It is easily (indeed, I'd say more easily) parsed as an introduction to a gene that codes for the trait of selfishness.

Whatever faults you might have with Schneier's title, I don't think you can say that there are multiple ways to parse it.


I agree that these are two ways to parse "The Selfish Gene" and the first one that you listed is more accurate, but what I don't like about the title is that is sounds a bit like an excuse that someone would use.

"Hey, that's just the way my genes made me : I'm selfish, okay, but that's, like, genetics, survival of the fittest so it's all fine and I don't have to feel bad about it ..."

The bad thing about the "Dishonest Minority" title is that is sounds like he's referring to a bunch of freeloaders who are abusing the system and ruining it for every honest tax-payer. From reading his excerpt, we know that it's not what he's saying at all. I like the shock value of using "Dishonest Minority", but sometimes book with titles like that nuance it by having a subtitle that goes like "How we all benefit from them" or something like that.


    All complex systems contain parasites. In any system of cooperative
    behavior, an uncooperative strategy will be effective -- and the system
    will tolerate the uncooperatives -- as long as they're not too numerous
    or too effective. Thus, as a species evolves cooperative behavior, it
    also evolves a dishonest minority that takes advantage of the honest
    majority. If individuals within a species have the ability to switch
    strategies, the dishonest minority will never be reduced to zero. As a
    result, the species simultaneously evolves two things: 1) security
    systems to protect itself from this dishonest minority, and 2) deception
    systems to successfully be parasitic.
That would be a pretty good description of IP. IP is in economic terms non-rival and non-excludable. It turns out that a free market does not work for goods that are non-rival and non-excludable (by "work" I mean reach a theoretical optimal allocation of resources to production). We developed a cooperative behavior in order to make it so a free market would work for such goods--basically by PRETENDING that IP is rival and excludable.

Some chose not to cooperate, and ignore that. As long as they are not too numerous, the system as a whole does not break down. The majority pretends that IP is rival and excludable, so pay creators for their creations, and creators get paid roughly in proportion to the demand for their works, and thus creators tend to create the works that the consumers want.

DRM has been developed to attempt to limit the effects of the uncooperative, and the uncooperative have developed systems to try to make it harder to detect piracy.


I would argue that it's a subset of the IP holders who are the parasites here, using their wealth to increase their rights beyond what is reasonable and in the interests of those with whom they are supposed to be cooperating. Piracy is the reaction Schneier mentions to the noncooperatives becoming too numerous and too effective. DRM is yet another strategy to limit the rights of those who continue to cooperate.


There is more than one kind of parasite.

Copyright and other intellectual exclusion (IE) laws were introduced in good faith to ensure that (1) it is rational to undertake intellectual endeavours and (2) the number of people who benefit from intellectual endeavours but don't support its development is limited.

Then the background against which IE had been written changed; now, some IE holders began to abuse their monopoly (a form of parasitic behaviour), and they came to informal understandings with lawmakers (who also put their own personal interests ahead of the collective interest, which is again parasitic) to strengthen the IE laws.

The real problem with IE laws is that as more people use intellectual works, the value to society increases, but due to the exclusionary nature of IE laws, they discourage widespread dissemination. The best solution would be a separate parallel economy for intellectual works where all consumers pay an amount from the primary economy that doesn't depend on how many intellectual works they consume, but producers receive an amount that depends on how much utility consumers directly or indirectly extract from their work (in proportion to all other intellectual works). That would open a number of difficult issues, and current parasites under the IE scheme have amassed enough power to resist it strongly, so I doubt we would see it any time soon even though it would be collectively beneficial.


It is also a little more complicated than that. The corporations are themselves uncooperative. They have broken the system by extending it beyond its basic intent for increased profit. They further seek to use the spectre of piracy to give them greater profit levels. Such as the establishment of taxes on CDRs in some nations which is given to the supposed losers.


Unfortunately, the dishonest minority has learned to disguise itself and move into enforcing things that benefit the state at the expense of the populations it feeds upon. With democracy and a lot of influence by government over the media, it can often even dupe those populations into thinking that the population is in control and are the beneficiaries.


While not related specifically to security, I've always been fascinated with the fact that Japan recognizes their 'dishonesty minority' the yakuza as legitimate organizations. While most countries run around declaring war on their black market parasites, Japan has allowed them to slice of their piece of the pie and avoid the nastiness.


That's not unique to Japan, criminal organizations have been able to gain differing levels of legitimacy and tolerance in just about every society, from the current Russian government to the American mafia in Nevada to too many examples of corruption around the world to list. And no, it is not commendable.


It's interesting to jump between his opening comment on society 'tolerating' the uncooperatives (as long as they're not too numerous or too effective) to his end-point that the dishonest minority serves as a catalyst for social change.

His statement of toleration of uncooperatives doesn`t seem to capture that society is better off with some of them (the dissenters) and that its actually in the interests of the governors to tolerate a certain amount of uncooperatives - not just because it overly onerous not to, but because otherwise stagnation will make the whole society vulnerable to an outside competitor.


His statement of toleration wasn't meant to illustrate the fact that society is better off with some dissenters. The two facts:

   * Society tolerates a certain level of dissension

   * Society is better off with a certain level of dissension
are not necessarily connected, other than their focus, and one does not necessarily imply the other. They can be viewed as two distinct facets of the phenomenon he is exploring.


All I know about publishing is from reading Charles Stross' blog (especially http://www.antipope.org/charlie/blog-static/2010/04/common-m...), but doesn't manuscript-submit in November 2011 and publishing in February 2012 seem awfully fast?


That's only the final manuscript---he's got a draft now and the editing process will be ongoing well before November.


It also seems very slow. Why does proofreading and typesetting take more than 2-3 days?


Manuscripts come filled with errors, and it takes a lot of time to find all of them and correct them. Errors include punctuation, spelling, grammar, word choice and even style. It's not just something you can leave to a speedreader or a computer program.

Typesetting is the precursor to (web) page design. There's a lot of work involved. Programs like Adobe inDesign streamline the tedious aspects of page design, but not the work that goes into actually designing the look of each page.


Bruce seems to be associating dissent with sabotage here: Vibrant societies need a dishonest minority; if society makes its dishonest minority too small, it stifles dissent as well as common crime.

I disagree with this vehemently. Dissent is not crime.

And his argument that society needs a bit of dishonesty is also a bit odd. I see it as something any large cooperative system is doomed to end up with. This is also what his first paragraphs explains. And I think there have been software simulations which show how both "criminals" and "cops" naturally arise in a complex cooperative system.

I think his attempt to link dissent with crime and dishonesty is a terrible way to defend the need for dissent.


Dissent IS crime though. For example, stealing is dissent from the accepted notion that property can be owned. It's an assumption so baked into your brain that you accept it implicitly, but it's not a law of physics or anything.

Any deviation from the rules of the system is a challenge to those rules. I think the distinction between crime and dissent in your mind is actually just whether you agree with the person doing it or not.

Civil Rights activists were criminals. Until they weren't.


Dissent IS crime though. For example, stealing is dissent from the accepted notion that property can be owned.

That proves crime is (in some loose sense) dissent, not that dissent is crime.

Besides, stealing generally doesn't come from a dissent about the nature of ownership. The average thief doesn't believe that property can't be owned -- he knows that you own your property, and he intends to make it his property.

Any deviation from the rules of the system is a challenge to those rules. I think the distinction between crime and dissent in your mind is actually just whether you agree with the person doing it or not.

Well no, because:

a) "Rules" != laws" -- breaking "rules" isn't a crime while breaking laws is, and

b) You can quite easily disagree with the law without breaking it. I think it should be legal to serve alcohol to eighteen-year-olds in the state of Kansas, but I've never done it.


Merely disagreeing with a law doesn't lead to eventual change. You have to break it more and more visibly until they prosecute you, and then get enough people to rally around you and win the case, and the appeal, etc.


Merely disagreeing with a law doesn't lead to eventual change. You have to break it more and more visibly until they prosecute you, and then get enough people to rally around you and win the case, and the appeal, etc.

No, that's one way of changing a law, and I generally don't think it's the best. The best way is to go round persuading people of your case.

If you declare that it's morally alright for folks to break laws just because they happen to disagree with them, then the biggest problem you encounter is that folks' opinions on what laws are unjust are often just plain wrong. For instance, if Fred Bloggs believes that it should be legal to beat up queers, then he's perfectly entitled to that opinion, and to try to use his free-speech rights to persuade others that we ought to change the law so that it's legal to beat up queers. But it doesn't make it on some level okay for him to start beating up queers.

Or a more morally neutral example: some people think we should drive on the left, some people think we should drive on the right. But civil disobedience is really bad way to sort that one out.


What do you propose for a group of people who have little to no political power, and are being oppressed by some current law? They can run around telling people their argument all day long, but it is protests, and civil disobedience that have always forced those in power to address their grievances.

In regards to you gay-bashing example, civil disobedience is by definition non-violent. If a law was passed that required people to give gay people high fives, and Fred Bloggs chose to refrain, THAT would be civil disobedience.


I have to disagree with this definition, because it doesn't have to be breaking a law. It just has to be breaking a 'rule'. Breaking a rule doesn't necessarily equate crime.

As an example, it may be a rule to be on the job by 8 am. Being late can get you fired, but in most cases it's not going to land you in jail because being late to work is not illegal.


And how did we get to this point from the word "dishonest", which is what we're talking about here?


Bruce isn't saying that dissent is a crime, he's saying that the extreme measures we would need to use to reduce dishonesty (good) would also have the negative impact of stifling dissent (very bad) - so we shouldn't use these extreme measures.


Dissent is the expression of opposing opinions from the law. This often takes the form of crime.

When Rosa Parks sat in the front of the bus, she was expressing dissent. But she was also violating the law, and therefore committing a crime.


Based on the quality of the author's writing and the clarity of his thinking, I'm looking forward to reading that book. His book Applied Cryptography was one of the best written CS books I've ever seen.





Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: