home : about : p2p : security : technology : trust
Total entries: 72.
This has become a bibliography and comment section, and is most useful when looked at in conjunction with the reading list of books I intend to get to reading at some point. I think that it's useful to show people may current background and state of thought, and this is probably the easiest way of doing so. I don't pretend to have produced a useful critical outline of each resource: that's not what this section is about. The intention is to allow me to record my thoughts about resources, and only, really, in terms which I think are likely to be relevant to my field of study - trust relationships in the distributed peer-to-peer (p2p) world. Neither do I pretend to a consistency of classification - the trust section, in particular, might be better expressed as a subsection of the sociology section. I feel it's useful to separate out certain aspects, however, and expect to continue doing so. I'm also aware that a number of anthropologists and animal psychologists might be unhappy with my lumping of their disciplines together, but it works for my purposes, and I have no plans to change it in the near term.
I've listed articles within books separately, rather than under a general heading, as this is how I'm likely to reference them. I'm aware that this section has got rather long (partly due to this decision), and I've given some thoughts about how to remedy this, but have come to the conclusion that other than splitting it into sections (which I've done), it's not actually worth separating out things much more. I suspect that this page will act as a better resource for people to scan over my thoughts - maybe responding to them (firstname.lastname@example.org) - if the "complete set" is there for all to see. At some point, I may have to split out the sections, but I suspect that the sociology section is going to remain the largest, so it's probably not worth it. Comments are welcome.
Harvard reference style (or something resembling it) is used throughout.
- Andrew Whiten and Richard W. Byrne - The Machiavellian intelligence hypotheses: editorial A brief overview of the background to the hypothesis/es that social (Machiavellian) intelligence is not only separate from "technical" intelligence, but also that it was a major driver in human and primate evolution - without it, we would not be as intelligent as we are in any sphere. (Whiten, A. & Byrne, R.W. "The Machiavellian intelligence hypotheses: editorial" in Byrne, R.W. & Whiten, A. (eds) (1988) Machiavellian intelligence: social expertise and the evolution of intellect in monkeys, apes and humans, Clarendon Press, Oxford, pp. 1-9)
- Robin Dunbar - Grooming, gossip and the evolution of language - a rather "popular science" feel, despite the long bibliography. Maybe it was this approach to the book which led me to become rather annoyed by the the lack of scientific rigour to many of Dunbar's suggestions as to why language has developed as it has. His basic view is that as the size of human groups rose to around 150, we learnt to talk so that we could "groom" more people than we would be able by physical contact alone. A feasible hypothesis, but I was far from convinced that his rather flimsy set of proofs presented a stronger argument than some of the others against which he argued. I'm not sure if this was due to the expected audience, or a failing of rigour. One point, however, is important to the area of concern - that human groups max out at around 150, and that we require social contact to keep them working to that extent. The extent to which this can be provided without actual real-life contact is not addressed - the book was only written in 1996, and he is fairly gloomy about the prospect for online community in the brief mention he makes of it (he's basing his prognosis on email contact only, as well). He cites Enquist and Leimar on controlling free-riders "by making the formation of a coalition a costly business, by individuals demanding some token of commitment before agreeing to become involved." (p.45) (Dunbar, Robin (1996) Grooming, gossip and the evolution of language, Faber and Faber, London)
- Nicholas K. Humphrey - The social function of intellect One of the early works in the "social intelligence" canon, Humphrey argues that evolution ismean with its favours, and that social intelligence would not have evolved if it were not of use.
As he points out, there will be limits, for instance "where the time to resolve a "social" argument becomes insupportable." (p.21) (Humphrey, N.K. (1976) "The social function of intellect" in Byrne, R.W. & Whiten, A. (eds) (1988) Machiavellian intelligence: social expertise and the evolution of intellect in monkeys, apes and humans, Clarendon Press, Oxford, pp. 13-26)
- "...[I]n a society of the kind outlined [one where the learning of "subsistence technology" by young members of a group takes place in a "collegiate community" of considerable strife], an animals' intellectual "adversaries" are members of its own breeding community. If intellectual prowess is correlated with social success, and if social success means high biological fitness, then any heritable trait which increases the ability of an individual to outwit his fellows will soon spread through the gene pool. In these circumstances there can be no going back; an evolutionary "ratchet" has been set up, acting like a self-winding watch to increase the general intellectual standing of the species." (p. 21)
- Alison Jolly - Lemur social behaviour and primate intelligence Jolly contrasts learning about objects and social learning, pointing out that lemurs typically pick up object-related tasks more quickly when there is a social aspect to them. (Jolly, Alison (1966) "Lemur social behaviour and primate intelligence" in Byrne, R.W. & Whiten, A. (eds) (1988) Machiavellian intelligence: social expertise and the evolution of intellect in monkeys, apes and humans, Clarendon Press, Oxford, pp. 27-33)
- Michael J. Fischer - A Theoretician's View of Fault Tolerant Computing A general introduction to fault tolerant systems, including terminology and some of the major issues. Two important definitions:
A number of trust interactions may, it seems to me, be modelled as systems with possible Byzantine failures. (Fischer, M. J. (1987) "A Theoretician's View of Fault Tolerant Computing", in Simons, B. & Spector, A. (eds) Fault Tolerant Distributed Computing, New York, Springer-Verlag, pp. 1-9)
- Failstop fault A faulty process ceases operation and other processes are notified of the fault
- Byzantine fault A faulty process may continue to operate but can send arbitrary messages
- Ian Foster & Adriana Iamnitchi - On Death, Taxes, and the Convergence of Peer-to-Peer and Grid Computing Good discusion of the differences and similarities between P2P and grid computing, suggesting a likely convergence. Their definition of the two technologies is useful:
They also characterize them thus: "Grid computing addresses infrastructure but not yet failure, whereas P2P addresses failure but not yet infrastructure". A useful primer, with a good set of references. (Foster, I. & Iamnitchi, A. (2003) On Death, Taxes and the Convergence of Peer-to-Peer and Grid Computing, online at <URL: http://iptps03.cs.berkeley.edu/final-papers/death_taxes.pdf>, accessed 2003-04-12)
- "...Grids are sharing environments implemented via the deployment of a persistent, standards-based service infrastructure that supports the creation of, and resource sharing within, distributed communities."
- "We define P2P as a class of applications that takes advantage of resources - storage, cycles, content, human presence - available at the edges of the Internet."
- Jim Gray - A Comparison of the Byzantine Agreement Problem and the Transaction Commit Problem States that:
(p.14) I'm not sure from context whether this is for synchronous or asychronous systems, but I believe synchronous. (Gray, J. (1987) "A Comparison of the Byzantine Agreement Problem and the Transaction Commit Problem", in Simons, B. & Spector, A. (eds) Fault Tolerant Distributed Computing, New York, Springer-Verlag, pp. 10-17)
- If at least 1/3 of the generals (processes/entities) are bad, then the good generals cannot reliably agree
- If fewer than 1/3 of the generals are bad, then there are many algorithms [to reach agreement]
- Solutions to the algorithm have polynomial cost (e.g. ~N^2 messages) and, assuming constant service time for a broadcast, have constant delay (typically ~N^3)
- Fred B. Schneider - The State Machine Approach: A Tutorial An introduction to the use of state machines for fault tolerant systems analysis.
A formal approach to agreement between processes is presented. He also discusses reconfiguration and reintegration of repaired object, neither of which are particularly relevant in this case. (Schneider, F. B. (1987) "The State Machine Approach: A Tutorial", in Simons, B. & Spector, A. (eds) Fault Tolerant Distributed Computing, New York, Springer-Verlag, pp. 18-41)
- "When processors can experience Byzantine failures, an ensemble implementing a f fault-tolerant machine must have at least 2f+1 copies, and the output of the ensemble is the output produced by the majority of the state machine copies. This is because with 2f+1 copies, the majority of the outputs remain correct even after as many as f failures" (p.23)
- Pirsig - Zen and the Art of Motorcycle Maintenance. An interesting, first-person, discussion of the need to combine both the theoretical/formal and emotional/spiritual parts of our personalities if we are to be truly human.
- Melissa Scott - Trouble and her friends. Very believable community of practice portrayed - and an awareness of the nexus of communities of practice that we have. The trust exhibited by those who have met offline is clear, and also of those who are alike in other ways. In the main character's (Trouble's) community, it is those who have installed the (semi-legal) brainworm who are one community of practice, and a subset of them who identify themselves as "queer" (gay and lesbian), and are identified by others as such, who enjoy particularly high levels of trust. (Scott, Melissa (1994) Trouble and her friends, Tor, New York)
- John Casti - The Theory of Networks An introduction to networks that made me realise how long it was since I'd done any mathematics. Deals with Path Theorem, routes and optimal routes around a network, Eurlerian circuits (whether it is possible to visit each vertex in the graph and return to the starting vertex, traverse each edge only once, and Hamiltonian circuits (traversing an edge more than once). Also flows in networks, introducing cuts and the Min Cut-Max Flow Theorem.. Also optimal routing problems, weighted digraphs and pule processes. (Casti, John "The Theory of Networks" in Batten, D., Casti, J. and Thord, R. (eds) (1995) Networks in Action, Springer-Verlag, Berlin, pp. 3-24).
- Daniel Kahneman and Amos Tversky - Intuitive prediction: Biases and corrective procedures - this only just fits in the "mathematics" section, but there we go. A good suggestion for a procedure to correct errors in "intuitive prediction". As well as having a better awareness of distributive data (type data) as well as singular data (instance data) and an understanding of regressive effects, Kahneman and Tversky suggest the following steps to reduce errors: 1. Selection of a reference class; 2. Assessment of the distribution for the reference class; 3. Intuitive estimation; 4. Assessment of predictability; 5. Correction of the intuitive estimate. They give the following explanation and example:
(Kahneman, D. & Tversky, A. "Intuitive prediction: Biases and corrective procedures" in Kahneman, D., Slovic, P. & Tversky, A. (1982) Judgment under uncertainty: Heuristics and biases, Cambridge University Press, Cambridge, pp.414-420)
- "An estimate of the ordinal correlation between predictions and outcomes can ... be obtained as follows: If r is the estimated proportion of pairs in which the order of outcomes was correctly predicted, then t=2r-1 provides an index of predictive accuracy, which ranges from zero when predictions are at chance level to unity when predictions are perfectly accurate." (p.420, "r" and "t" substituted for Greek letters for HTML display)
- "...[S]uppose that the expert's intuitive prediction of the sales of a given book is 12,000 and that, on average, books in that category sell 4,000 copies. Suppose further that the expert believes that he would correctly order pairs of manuscripts by their future sales on 80 percent of comparisons. In this case, t= 1.6 - 1 = 0.6, and the regressed estimate of sales would be 4,000 + 0.6(12,000 - 4,000) = 8,800." (p. 420, "t" substituted for Greek letters for HTML display)
- Donald MacKenzie - Mechanizing proof: computing, risk, and trust - 0-262-13393-8. Not really about trust in the context in which I'm interested in the subject, but more about whether software (or more properly, software-hardware systems) can be "proved" to be correct, how this might be attempted, some of the issues, and some of the ramifications. One thing it did open my eyes to is the relevance of fault tolerant systems to the whole issue of p2p trust. (MacKenzie, Donald A. (2001) Mechanizing proof: computing, risk and trust, MIT Press, Cambridge, Mass.)
- Peer to Peer from O'Reilly - ISBN 0-596-00110-X. An interesting introduction, lots of stories, and some theory. In particular, the chapter on "Trust" was more about dealing with the fact that most users are distrusted - the chapter on "Reputations" was more useful, though very specific to a particular product.
- Nicholas Carroll - Spinning the Web - the Realities of Online Reputation Management Carroll examines the creation and management of reputation - particularly organisational reputation - on the Web. He concludes that it is more easy to demolish a reputation on the web than it is to build one: "Outside of e-commerce, the Web is presently a fairly weak means of enhancing one's reputation or agenda", though he is hopeful for the future, and "the convergence of opinion". (Carroll, Nicholas, Spinning the Web - the Realities of Online Reputation Management, online at <URL: http://www.mindjack.com/feature/spin.html>, accessed 2003-04-13).
- Ryo Chijiiwa - Reputation Economy and the Internet In this short article, Chijiiwa discusses the relevance of reputation as a value system on the Internet, concluding that although it has its problems, "it generally provides a more even playing field that it more likely to reward members based on their true capabilities than is typical in a monetary economy." The article raises issues of the timeliness of reputation, and of how marketing, brand and lobbying can influence reputation. The accent is on the Internet as a community of "hackers" (in the positive sense, rather than as "crackers"), particularly with regard to Open Source projects and is therefore looking mainly at a subset of the Internet as a whole. (Chijiiwa, Ryo, "Reputation Economy and the Internet", online at <URL: http://ryo.iloha.net/?section=writing&page=reputation_econ>, accessed 2003-04-14)
- CmdrTaco - Slashdot Moderation - guidelines for moderation of articles on the slashdot news service. (CmdrTaco (1999), Slashdot Moderation, online at <URL: http://slashdot.org/moderation.shtml>, accessed 2003-10-05)
- Eric J. Friedman and Paul Resnick - The Social Cost of Cheap Pseudonyms An examination of how reputation systems (focusing on eBay) can work where it is relatively cheap to change pseudonyms. They suggest a few ways to manage the problem:
An excellent use of game theory to attack this difficult problem. (Friedman, E.J. & Resnick, P. (1999) The Social Cost of Cheap Pseudonyms online at <URL: http://www.orie.cornell.edu/~friedman/pfiles/anon.pdf>, accessed 2003-09-27)
- distrust newcomers: costly to the system
- increase the cost of name-changes: impacts on poorer users, and has ramifications on who would change names
- have an external trusted third party who issues one-off name-changes for over a specified game period: has problems with what constitutes a "game" period, and with election of a trusted third-party.
- Ross Anderson - Security Engineering: A Guide to Building Dependable Distributed Systems - ISBN 0-471-38922-6. A fascinating introduction to many areas of security engineering and how they interact in distributed systems. Some useful thoughts at the end about assurance and evaluation, as well as engineering methods. Illustrated with the stories for which Anderson is well known, it's a must read for anyone looking for a wider view of secure systems, and Anderson would say that anyone designing a secure system should know more about other aspects of security. Heartily recommended, though not short with over 530 pages. (Anderson, Ross (2001) Security Engineering: A Guide to Building Dependable Distributed Systems, Wiley, New York)
- James Essinger - Internet trust and security: the way ahead - ISBN - 0-201-72589-4. Actually not very relevant: it's a business-oriented examination of of some of the issues to do with e-commerce and e-business, leaning heavily on a report by Ovum in 2000. (Essinger, James (2001) Internet Trust and security: the way ahead, Pearson Education, Harlow)
- Charles P. Pfleger - Security in computing - a standard introduction to security in computer systems, this provides a solid grounding and overview of the subject. (Pfleger, Charles P. (1997), Security in computing, 2nd ed., Prentice Hall, Upper Saddle River)
- Ronald Rivest - chaffing and winnowing - a great technique for hiding information in a data stream without using cryptography. I've written a basic implementation. (Rivest, R.L. (1998), Chaffing and Winnowing: Confidentiality without Encryption, MIT, online at <URL: http://theory.lcs.mit.edu/~rivest/chaffing.txt>, accessed 2003-04-14)
- Bruce Schneier - Beyond Fear: Thinking Sensibly About Security in an Uncertain World - an attempt to present the area of security (and not just technical security) to a broad audience and to demystify it. Schneier presents a five-step process to analyze the trade-offs made in any security decision, and gives examples. The steps are:
As always, Schneier provides concise and cogent arguments, and the book is very successful. It is aimed at a post 9/11 world, and there's a definite slant towards helping (U.S.) Americans to make judgements about the need for and the success of the security processes put in place in their country since the World Trade Center bombings (he argues that the great majority of them are unnecessary). However, he is careful to keep the book's appeal broader than this, and provides a host of international examples. There is a great deal of discussion of the importance of resilience in security systems through avoiding "brittleness" and of how it's very preferable for a system to fail securely, rather than insecurely (e.g. if a nuclear power station has a problem, it should shut down, not be allowed to go into meltdown). He has a good section on trust (from which a few brief quotes):
- Step 1: What assets are you trying to protect?
- Step 2: What are the risks to these assets?
- Step 3: How well does the security solution mitigate those risks?
- Step 4: What other risks does the security solution cause?
- Step 5: What costs and trade-offs does the security solution impose?
There are also good sections on trusted third parties, the importance of protocols (outside the technical sphere) and the difficulties of getting people to follow them, particularly if they're rarely practised or required. (Schneier, Bruce (2003) Beyond Fear: Thinking Sensibly About Security in an Uncertain World, Copernicus Books, New York)
"There are three basic ways to secure trusted people - or trust machines, for that matter. The first: Try to put trustworthy people in positions of trust, and try extra hard for positions of extreme trust. This is why background chekcs or drug testing are done on key employees." (p. 139)
"The second way to secure trusted people is through compartmentalization. Trust is not black and white, meaning that someone is either trusted or not. And trust is not monolithic, meaning that trust people are not all trusted to the same degree and in the same contexts. We can increase security by limiting our trust: Give trusted people only the information, access, and compatibilities they need to accomplish their tasks." (p. 140)
"Trust should not only be limited, it also should be granted directly. If you trust Alice, and Alice trusts Bob, that does not automatically mean that you trust Bob. You might not even know Bob. In systems where you automatically trust Bob, trust is said to be transitive. And when trust must be transitive, it creates brittleness." (p. 141)
"The third way to secure systems with trusted people is to apply the principle of defense in dpeht: Give trusted people overlapping spheres of trust, so they effectively watch each other." (p.141)
- Bruce Schneier - Secrets and lies: digital security in a networked world - a fantastic introduction to why cryptography doesn't solve the world's ills (despite Schneier's earlier thesis that it would). (Schneier, Bruce (2000) Secrets and lies: digital security in a networked world, John Wiley & Sons, New York)
- Joe Stewart - DNS Cache Poisoning - The Next Generation - an examination of some of the the well-known problems with DNS protocol, and the latest attacks on it from DNS cache poisoning. This is interesting in terms of trust because DNS is an example of a component of a system where a large number of entities trust a small number of other entities. (Stewart, Joe (2003) DNS Cache Poisoning - The Next Generation, online at <URL: http://www.securityfocus.com/guest/17905>, accessed 2003-10-05)
- Benedict Anderson - Imagined Communities - ISBN 0-860-91546-8. Not what I'd expected (a look at non-physically-based communities such as on the Internet), but a look at the growth of the idea of nation, and the pre-requisites for its development (such as print capitalism). A little light on theory, and what was there seemed overly-reliant on empirical evidence from the Anderson's earlier work in the East (Philippines, etc.).
- Robert Axelrod - The Evolution of Co-operation - ISBN 014-012495-0. A discussion of how, in a long running set of interactions of Prisoner's Dilemma games, given appropriate conditions (such as the a high "shadow of the future" - a specific range of discount factors), cooperation is a strong strategy for individuals and clusters of individuals in a population, whether this population is made up of forward-thinking individuals (such as units in WWI) or not (such as bacteria in a host). Axelrod suggests the following rules for a successful strategy:
To increase the chances of cooperation in an environment, he suggest the the following changes to an environment:
- Don't be envious
- Don't be the first to defect
- Reciprocate both cooperation and defection
- Don't be too clever.
He points out the importance of reputation and deterrence and the balances that a government must make between allowing flexibility and forcing compliance. He also suggests that territoriality is an important factor, as it forces repeated interactions with known and easily recognised individuals. A useful book - though less academically rigorous than it might be (due to its intended popular science market, I assume). (Axelrod, Robert (1990) The Evolution of Co-operation, Penguin Group, London).
- Enlarge the shadow of the future
- Change the payoffs
- Teach people to care about each other
- Teach reciprocity
- Improve recognition abilities.
- William Bogard - The simulation of surveillance - ISBN 0-521-55561-2 (paperback), 0-521-55081-5 (hardback). A masterly examination - deeply rooted in Baudrillard and Foucault - of the possibilities of how our futures as "cyborgs" - parts of the telematic/informational world - may play out. Bogard takes a postmodern - and very post-structuralist - approach, arguing that surveillance becomes a nonsense once we are all part of the thing being surveilled - we are part of the simulation. The difference between us and the information that represents us withers away, and our identities are as much "online" as they are in "the real world": we have become "cyborgs". In an issue related to my specific interests, he paraphrases Steven Nock thus:
I have a lot of sympathy with this view, although I follow Bogard in mistrusting Nock's view that surveillance can restore trust as it "fills the gap created by an explosion of privacy", not necessarily for the same reasons as Bogard, but because it become senseless to trust the information that is provided by "surveillance" as it is not more trustworthy than the information we are trying to replace. Another quote from Bogard (taken slightly out of context, as he is talking about gender relations in particular) helps to show his basic thesis:
- "He claims that the historical emancipation of young people from traditional family structures has in fact expanded the modern domain of privacy and produced in its wake a radical crisis of trust. Too much privacy, and no one knows anyone else intimately enough to inspire any confidence in social relations." (pp. 147-148)
(Bogard, William (1996) The simulation of surveillance: Hypercontrol in telematic societies, Cambridge University Press, Cambridge)
- "It is less helpful [than looking beyond gender to the medium itself] to think of men and women "using" or "communicating over" or "being exploited by" these systems than as virtual nodes or switches in cyberspace and cybertime, whose gender is more intimately related to the logic and design of the technology than to any "real" social context of the operator. The "social," if there still is one here, in on [of? - MHMB] the Net." (p. 177)
- Ronald L. Breiger - The duality of persons and groups A look at how networks of interpersonal relationships and group relationships can be expressed in matrix form and a person-to-group affiliation matrix combining the two can be used to look at cliques and other types of groups. Not an introduction, but interesting to see what tools are available. (Brieger, Ronald L. "The duality of persons and groups", in Wellman, B., and Berkowitz, S.D. (1988) Social Structures: A Network Approach, Cambridge University Press, Cambridge, pp. 83-98)
- Rupert Brown - Group Processes: Dynamics within and between groups - ISBN - 0-631-21852-1. A discussion of groups from a social psychology perspective (rather than a purely sociological one), with a survey and discussion of many of the different theories in the field. Brown breaks the subject down into eight chapters:
Though issues such as trust are not explicitly tackled in the book, a number of the subjects surveyed are of interests in terms of how to increase performance in a group. In particular, the importance of task goals and of cohesion as commitment to group goals, are rather salient. Questions of categorization may also prove to be of interest. (Brown, R. (2000) Group Processes: Dynamics within and between groups, 2nd edition, Blackwell Publishers, Oxford)
- The reality of groups - definition of groups for the purposes of the book; discussion of whether a group is more than the sum of its parts and their behaviour; group behaviour as more uniform than individuals, and more stereotyped; crowds can be pro-social; intergroup and intragroup relations interesting, and social identity versus personal identity often important
- Elementary proccesses in groups - pre-joining reconnaissance to assess rewards and costs; membership can have positive and negative effects on self-esteem; initiation often marked by ritual, delineating members from non-members, initiation discomfort may increase commitment; binding of ones fate with others; task goals very important for formation; behaviour focussed on group goals and behaviour around feelings (socio-emotional); cohesion as attraction to ideal or prototype; cohesion as commitment to group goals; cohesion leading to increased adherence to group goals, but not always enhanced group performance; norms defining (un)acceptable behaviours.
- Structural aspects of groups - role and status key aspects of group structure; hierarchies form from power and prestige differences, leading to (often self-fulfilling) expectations of behaviour; social comparisons of abilities, often to those similar; "traits for leadership" not backed by evidence, leaders need attention to task and consideration for members; good leaders have personalities matched to particular situations; leadership legitimacy acquired in various ways; network of communication in a group importance, topological views of networking over distance theories, decentralized systems better for complex problems.
- Social influence in groups - conformity to attitude and behaviour of majority widespread; why? - for information about the world and validity of own views, for achievement of group goals (easier with uniformity of purpose), for need for approval through seeming similar; deviates from uniformity attract most attention from majority and are liked less; deviates can influence majority if consistent, particularly where measures are indirect or latent - ingroup minorities more so than outgroup minorities; unclear whether minority and majority influences arise from different processes.
- Individuals versus groups - performance affected by presence of another person, positively for simple tasks, negatively for complex ones; interacting groups can perform less well than sum of their members; may be due to underutilization of resources (Steiner) or lack of motivation (Latané); groups can lead to gains, particularly for complex tasks, where group is psychologically important to members, or prevailing values favour collectivism; social decision schemes theory as a way of studying group decision making; group decisions on judgemental tasks often polarized; polarization theories - social comparison theory, persuasive arguments theory and social identity theory; groupthink (Janis) can arise where groups don't consider all relevant information and/or appraise full range of options.
- Intergroup conflict and cooperation - intergroup prejudice and social discontent linked by frustration-aggression theory; relative deprevation theory as a development; conflictual goals lead to antagonism; superordinate goals lead to friendliness and cooperation.
- Thinking about groups - categorization of the world key to human cognition; use of one categorization over another depends on situation; categorization can lead to behaviourial discrimination, and bias towards ingroup; stereotyping as attribution of common characteristics of a group; three important aspects of stereotypes - legitimating existing intergroup inequalities, expectancies that bias judgements; self-fulfilling qualities on those stereotyped.
- Social identity and intergroup relations - social identity theory providing explanation for many ingroup-favouring biases; most intergroup situations involve status and power differences; acquisition/maintenance of self-esteeem not the only reason for social identification; intergroup prejudice can be reduced by bringing groups together in conditions involving cooperation between equal-status participants; where category memberships overlap, shared gopur identities can exist, though one group identiy may dominate to exclusion of others; changing salience of group identities may allow attitude change - through decategorization, recategorization and categorization.
- James Coleman - Individual interests and collective action - ISBN 0 - 0-521-30347-8. A collection of essays, mainly around social welfare, but also on the cross-overs between political theory, economic theory and sociological theory. Of particular interest was the chapter "Inequality, sociology and moral philosophy", a criticism of John Rawls' "A Theory of Justice" (1971). The main point of interest is an attempted improvement over Rousseau's and Hobbes' contract theory, whereby "natural persons" establish implicitly or explicitly social contracts with multiple corporate actors (one of whom may be the state), rather than investing in a single sovereign state. (Coleman, J.S. (1986) Individual Interests and Collective Action: Selected Essays, Cambridge University Press, Cambridge.)
- Drucilla Cornelle - Beyond Accommodation: ethical feminism, deconstruction, and the law - ISBN 0-415-901057 (hardback), 0-415-901065 (paperback). An attempt (generally successful, I'd judge, given my limited grasp) to criticise various readings of feminism within the deconstructionist school, including Irigaray, West, Derrida, Kristeva, Gilligan, Willig and MacKinnon. Inasmuch as the law represents the objectification (my word) of society's attitude to woman and its privileging of a male discourse, even within a liberal ethics, Cornelle argues that it is possible to imagine a future encompassing equal men and women by exploring myths of the feminine, and without having to resort to an essentialist Woman. I found the underpinnings of Lacan's theories of the phallus a little difficult to take, but much of the theory resonated strongly with me. I read it to reacquaint myself with deconstructionist theory, and to see it applied in a sociological context. Some useful passages (in terms of deconstructionist theory):
- "Différance can be understood as the "truth" that "being" is presented in time and, therefore, there can be no all encompassing ontology of the "here" and "now." Any reality, including the Law of the Father, established in and through the symbolic, is always already divided against itself as soon as it is presented." (p. 108)
- ""Being" cannot be separated from "seeing," but it cannot be reduced to it either. We do not see what "is," directly. We see through the world presented in language. ... [T]his world is never just presented as static, because the very language which allows us to "see" also allows us to see differently, because of the perfomative power of the metaphors that constitute reality." (p. 131)
- William Davies - You Don't Know Me, but...: Social Capital & Social Software - a examination of whether ICTs are eroding (as per Robert Putnam) or building social capital, and whether "social software" (broadly defined as technologies such as email, weblogs and Wikis, messenger systems, document editing systems, group diaries, introducer systems and group discussion systems) have a role to play. Several useful points are made:
The report is social commentators and decision-makers, and comes form "the work foundation", a policy think-tank. (Davies, William (2003) You Don't Know Me, but...: Social Capital & Social Software online at <URL: http://www.theworkfoundation.com/pdf/1843730103.pdf>, accessed 2003-09-27)
- online networks tend to work best if they reflect "real-world" social networks
- offline networks may be better at expressing tacit knowledge, online networks at codified knowledge (although I don't agree with this in all cases)
- that anonymity is not necessarily the best condition for social interaction; that non-verbal communication is important in interactions; that moderation of communications may be useful
- small groups tend to be more coherent (though this isn't how the sentiment is expressed).
- Morton Deutsch and Robert M. Krauss - The effect of threat on interpersonal bargaining. A study on the differences in bargaining when partners can threaten each other or not (or unilaterally), and when communication is allowed or not (or unilaterally). They found that bilateral positions of threat were the most dangerous, and that communication made less of a difference than expected. (Deutsch, M. and Krauss, R. M. (1960) "The effect of threat on interpersonal bargaining", in Journal of Conflict Resolution 6, 52-76)
- Bonnie H. Erickson - The relational basis of attitudes. A very good explanation of how network analysis can be used to examine blocks, cliques and structural equivalence, mainly from the point of view of attitude similarity. Her examination of dyads and attitude similarity identifies four major indicators which are likely to affect attitude similarity: frequency of contact, multiplexity, strength and asymmetry. She also looks at spatial models, and their problems. Many of these areas seem to me to be very applicable to p2p trust issues, particularly the issue of asymmetry and spatial modelling (Erickson, Bonnie H., "The relational basis of attitudes", in Wellman, B., and Berkowitz, S.D. (1988) Social Structures: A Network Approach, Cambridge University Press, Cambridge, pp. 99-121)
- Francis Fukuyama - Trust: The Social Virtues and the Creation of Prosperity - ISBN 0-241-13376-9. Fukuyama examines the role of social capital, familism, individualism and socialibility in economic growth, in particular their effects on how corporations and companies form. He argues - strongly - that mutual trust, a commodity linked to (James Coleman's) social capital, is very important to how people in different cultures form groups and corporations. I feel that his definition of trust is not exact enough, and his central interest in economic and financial structures does not tally very closely with my interests in P2P networks, but his work is both interesting and very useful. A useful quote to illustrate his views for our purposes:
Fukuyama's accent on the different cultural practices of nations as diverse - and seemingly close - as America, Italy, Germany, Taiwan, Korea, China and Japan, and how they relate to family life highlights the problems that raise their heads for massively distributed p2p systems comprising members of different cultures. (Fukuyama, F. (1995) Trust: The Social Virtues and the Creation of Prosperity, Hamish Hamilton, London.)
- "...[a] society built entirely out of rational individuals who come together on the basis of a social contrract for the sake of the satisfaction of their wants cannot form a society that would be viable over any length of time. In a criticism frequently leveled at Hobbes, such a society can provide no motive for any citizen to risk his or her life in defense of the larger community, since the purpose of the community was [sic] to preserve the individual's life. More broadly, if individuals formed communities only on the basis of rational long-term self-interest, there would be little in the way of public spiritedness, self-sacrifice, pride, charity, or any of the other virtues that make communities livable. Indeed, one could hardly image a meaningful family life if families were essentially contracts between rational, sefl-interested individuals." (p.351 - a good criticism of economic liberalism and utility-maximisation theory follows.)
- Giddens (ed.) - Sociology: introductory readings - ISBN 0-7456-1873-1. A selection of introductory readings in sociology, which I zipped through to get a view of what an undergraduate should probably be expected to know and get some of the main issues in my head. I was surprised to see Saussure and structuralism given pride of place, and no post-structuralist input in the "theories" section. (Giddens, A. (ed.) (1997) Sociology: introductory readings, Polity Press, Cambridge.)
- Mark Granovetter - The Strength of Weak Ties: A Network Theory Revisited Granovetter examines work related to an earlier paper of his around the importance of weak ties within communities. His thesis, that too many strong ties can lead to cliques and stagnation, but weak ties provide for movement of ideas and trends around communities, seems generally supported by later work. It tallies well with Lave and Wenger's views on the importance of peripheral participation in communities of practice. (Granovetter, Mark "The Strength of Weak Ties: A Network Theory Revisited", in Marsden, P. V. and Lin, N. (eds) Social structure and network analysis, Sage Publications, Beverly Hills, pp. 105-130)
- Jurgen Habermas - Communication and evolution of society (translated with an introduction by Thomas McCarthy) - ISBN 0-7456-0846-9. What a way to get an introduction to modern sociology... My main problem is the underpinning of structualist theory that informs much of the work, with associated normative themes and lack of issue with an "objective" viewpoint which seems required for the historical materialist approach Habermas uses. The relating of theories of societal "evolution" and psychological development in terms of a Freudian-based view of the stages of the ego's growth also feel a little outdated. I need to look for a post-structuralist critique of Habermas' work.
- Friedrich A. Hayek - The Road to Serfdom - ISBN 0-226-32077-4. A strong apologia for what would now probably be called free market economics - what Hayek sees as a liberalism towards capitalism which supports individualism. Writing in 1944, he sees socialism as a precursor to fascism, and sees the dangers of socialist leanings all around him. A clear, if (now) clearly biased, defense of liberalism and free market capitalism. (Hayek, F.A. (1944) The Road to Serfdom, The University of Chicago Press, Chicago).
- Heap and Varoufakis - Game Theory, a critical introduction, ISBN 0-415-09402-X (hb), 0-415-09403-8 (pbk). Very good introduction to game theory, and its weaknesses in certain circumstances. All you need to know about game theory, including: Bayes's rule, CKR (common knowledge of rationality), CAB (consistent alignment of beliefs), NEMS (Nash equilibrium mixed strategies), SPNE (subgame perfect Nash equilibrium), Nash backward induction, Nash's bargaining solution, dominated strategies, ESS (evolutionary stable strategies) I'm pretty much in tune with their postscript: "The ambitious claim that game theory will provide a unified foundation for all social science seems misplaced to us. ... At root we suspect that the major problem is ... that people appear to be more complexly motivated than game theory's instrumental model allows and that a part of that greater complexity comes from their social location." A particularly good description of how the differences between a Hume-based approach and a Marx-based approach impact on how views of game theory may develop. The chapter on bargaining games is most likely to be relevant to my later deliberations, I suspect, thought the evolutionary game theory may well play strongly, too. (Heap, S.P.H. & Varoufakis, Y. (1995), Game Theory: A critical introduction, Routledge, London.)
- Hobbes - Leviathan, ISBN 0-521-56099-3 (hardback) 0-52156797-1 (paperback). I only read the first two parts: "Of MAN", and "Of COMMON-WEALTH", leaving "Of A CHRISTIAN COMMON-WEALTH". Hobbes central thesis is that men are, by default, in "a state of nature", in which they are brutish and and warre with one another. They create agree to give up some of their rights to take from each other (though never to defend themselves) by embuing rights into a sovereign body (pereferably, in Hobbes' view, a monarch), and this sovereign body has the power to create laws (which are, by definition, just), and exact punishment. The logic behind much of his reasoning is very strong, though there exist problems with issues of how such sovereign bodies are ordained (to pick a word quite carefully), and how they are sustained, if there is no continual re-establishment of the sovereignty by its body politic. This issue is less important to this issue of trust, where we can consider that such a continual re-esablishment might be feasible, but the tenet that there must exist a central, controlling, sovereign, is a strong one, and raises real issues to do with trust: can reputation ever be enough? The answer may well be "no", and herein lies much of the interest in the subject for me. (Hobbes, T., Leviathan, ed Tuck, R., (1996), Cambridge University Press, Cambridge.)
- Jean Lave and Etienne Wenger - Situated learning: Legitimate peripheral participation - ISBN 0-521-41308-7 (hardback), 0-521-42374-0 (hardback). A good description of how participants learn within communities of practice - and how they become members of that community at the same time.
Good, but not particularly relevant to issues of trust in the context of P2P systems. (Lave, J. and Wenger, E. (1991) Situated Learning: Legitimate Peripheral Participation, Cambridge University Press, Cambridge)
- "Legitimate peripheral participation refers both to the development of knowledgeably skilled identities in practice and to the reproduction and transformation of communities of practice." (p. 55)
- John Locke - An Essay Concerning Human Understanding (Abridged and Edited by A. S. Pringle-Pattison) - ISBN 0-391-00573-1. I read this as a felt I ought to get a better feeling for one of the important texts in the field of sociology (and for more insight into Hobbes, which it didn't really provide). Not very relevant, though interesting (and long) - I certainly feel better educated having read it. Like Hobbes, Locke feels that he has to bring God into much of the discussion, but it's interesting to note how little of the discussion actually hinges on a Deity. A very 1920's commentary made for some interesting glosses. One area which was interesting was Locke's views on how we should decide the probability of something being true, which, in certain circumstances, comes down to issues of trust in other people:
This seems to me a good set of issues around which to consider trust models! (Locke, John, (ed. Pringle-Pattison, A. S.) (1680/1924) An Essay concerning Human Understanding, reprint 1978, Harvester Press, Sussex)
- 4. Probability then being to supply the defect of our knowledge, the grounds of it are these two following:
First, the conformity of anything with our own knowledge, observation and experience.
Secondly, The testimony of others, vouching their observation and experience. In the testimony of others is to be considered: (1) The number. (2) The integrity. (3) The skill of the witnesses. (4) The design of the author, where it is a testimony out of a book cited. (5) The consistency of the parts and circumstances of the relation. (6) Contrary testimonies.
- Niklas Luhmann - Familiarity, Confidence, Trust. A rather self-consciously structuralist and philosophical approach to trust, compared to confidence, and how the two live in relation to a posited familiarity/unfamiliarity dichotomy (in which I was far from convinced). (Luhman, N. (1988) "Familiarity, Confidence, Trust" in Gambetta, Diego (ed.) Trust: making and breaking cooperative relations, Basil Blackwell, Oxford, pp. 94-107)
- P. V. Marsden and E. O. Laumann - "Collective Action in a Community Elite: Exchange, Influence Resources, and Issue Resolution". An study of the use Coleman's model of influence and power as like a free market system in an empirical study of a US town. Of particular note is their extension of an earlier model by Laumann and Pappi to examine resource bases of influence, which, with adaptation, may well be very useful.
Also interesting are a number of mathematical models used in the modelling process, which may be worthy of later study. They also identify a number of issues which question the perfect market-style Coleman model, including:
(Marsden, P.V. & Laumann, E. O. (1977) "Collective Action in a Community Elite: Exchange, Influence Resources, and Issue Resolution" in Liebert, R. J. & Imershein, A.W. (eds) Power, Paradigms and Community Research, Sage Publications, London)
- Resource leakage:
- resource slack:when participants do not apply their resources when issues are not of great enough importance to them
- free-riders: when participants do not apply their resources to an issue which is already very likely to fail or succeed, but hold them back for other issues in which they may have more sway.
- multiple systems: when participants are involved in multiple systems of action, they may choose to direct their resources to issues in particular fields of action where they have strong interests.
- System permeability: also related to multiple systems issues, this is when actions from those outside the system being examined (such as higher government authorities) impose decisions or when participants of the system expend resources outside the system under study.
- Multiple processes of issue resolution: "The Coleman model assumes a singular process of issue resolution in a system of action where exchanges are based on possession of generalized influence resources." (p. 239). This is not always realistic, for instance in the case where participants use resources in lobbying of an external party to influence a decision in the system under study.
- Onora O'Neill - A Question of Trust (The BBC Reith Lectures 2002) - ISBN 0-521-82304-8 (hardback), 0-521-52996-4 (paperback). O'Neill discusses the "crisis of trust" in public institutions, and how this can be managed, suggesting (among other issues) that transparency (a removal of secrecy) is not enough, if we cannot remove deception, and that true accountability involves the ability to make informed, deliberate (and individually autonomous) choices. A thought-provoking read. She also talks about the importance of "duties" in relation to "rights", arguing that you cannot really have one without the other - and that this is a fundamental underpinning of democracy.
- Thomas Paine - The Rights of Man - ISBN 1-84212-107-3. An astonishing attack on Old European government, as represented by monarchical rule, and a brilliant (and still relevant) advocacy and examination of representative democracy. He attacks the view (espoused by Hobbes and Paine's opponent, Burke) that a contract can be made with a sovereign power that does not need reaffirming, and holds up the (then new) democracies of France and America as models of decency, efficiency and correctness. He is a strong believer in free commerce - though against "the corporations". Reading him in the light of the later history of French Revolution, Napoleon, the American Civil War, McCarthyism and recent American history makes you wonder how right he really was, but I'm much closer to a position of republicanism (losing the monarchy) having read his arguments than I ever have been before. (Paine, Thomas (1791,1792; 2000) "The Rights of Man", in Common Sense & The Rights of Man, Phoenix Press, London, pp. 57-278)
- Howard Rheingold - The Virtual Community - ISBN 0-436-41214-4 (hardback), 0-436-20208-5 (paperback). I read this when it first came out, and it was one of the drivers that got me into online communities. It's dated, and misses some issues which have become important, but is generally very good. Rheingold looks at a variety of virtual communities and identifies a number of trends. He identifies, in particular, the addictive properties of virtual communities, and how real-life events (in particular births, deaths and marriages) seem to cement them. He is a strong believer in the growth of social capital and cohesion within virtual communities via informal and highly contextualised interaction (special interest groups from parenting to Grateful Dead fandom). (Rheingold, Thomas (1994) The Virtual Community: Finding Connection in a Computerized World, Secker & Warburg, London)
- Thomas Schelling - The Strategy of Conflict Firmly rooted in the Cold War (it was first published in 1960), Schelling examines game theory from a variety of angles, always trying to keep real life possibilities in mind. He examines hostage-taking, threats, promises, and similar issues in what must have been a landmark book for its time. It's still very much worth a read. Some useful points:
(Schelling, Thomas (1960) The Strategy of Conflict, Harvard University Press, Cambridge, Mass.)
- there are times when it's useful to reduce your capacity for action, as this will force your opponent to take a particular course of action (if I drop my wallet into a postbox, what's the point of a mugger attacking me?)
- Schelling is very strong on how tacit communication can be a very strong force (for instance, if two people are parachuted into the countryside, they may well be able to find each other - they'll head for a bridge or high ground, for instance). He says:
- "tacit agreements or agreements arrived at through partial or haphazard negotiation require terms that are qualitatively distinguishable from the alternatives and cannot simply be a matter of degree
- when agreement must be reached with incomplete communication, the participants must be ready to allow the situation itself to exercise substantial constraint over the outcome; specifically, a solution that discriminates against one party or the other or even involves "unnecessary" nuisance to both of them may be the only one on which their expectations can be coordinated." (p. 75)
- Amos Tversky and Daniel Kahnemann - Judgment under uncertainty: Heuristics and biases - an introduction to some of the mistakes that people commonly make when making judgments where rational judgment based on statistical understanding would yield better guesses or estimates. They identify Representativeness, in which probabilities are evaluated by the degree to which an event or object is representative of a particular class (Insensitivity to prior probability of outcomes; Insensitivity to sample size; Misconceptions of chance; Insensitivity to predictability; The illusion of validity; Misconceptions of regression), Availability, "in which people assess the frequency of a class or the probability of an even by the ease with which instances or occurrences can be brought to mind" (p. 11) (Biases due to the retrievability of instances; Biases due to the effectiveness of a search set; Biases of imaginability; Illusory correlation) and Adjustment and anchoring, in which "people make estimates by starting from an initial value that is adjusted to yield the final answer" (Insufficient adjustment; Biases on the evaluation of conjunctive and disjunctive events; Anchoring in the assessment of subjective probability distributions). Of these, two seem likely to affect trust models in terms of how people make judg(e)ments:
(Tversky, A. & Kahneman, D. "Judgment under uncertainty: Heuristics and biases" in Kahneman, D., Slovic, P. & Tversky, A. (1982) Judgment under uncertainty: Heuristics and biases, Cambridge University Press, Cambridge, pp. 3-20)
- Misconceptions of regression: regression suggests that if a particular sample is above a mean, then the next sample is likely to be below the mean. This can lead to misconceptions such that punishing a bad event is more effective (as it leads to a (supposedly causal) improvement) than rewarding a good event (as this leads to a (supposedly causal) deterioration. "Thus, the failure to understand the effect of regression leads one to overestimate the effectiveness of punishment and to underestimate the effectiveness of reward. ... Consequently, the human condition is such that, by chance alone, one is most often rewarded for punishing others and most often punished for rewarding them." (pp. 10-11). In a system where trust is important, and needs to be encouraged, trying to avoid this bias may be an important goal in the design of the system.
- Biases on the evaluation of conjunctive and disjunctive events: people are bad at realising that a number of unlikely, but conjunctive events (such as the chance of a broken chain of trust where all the links in the chain enjoy a high probability of trustworthiness) are likely to yield a bad outcome.
- Barry Wellman - Structural analysis: from method and metaphor to theory and substance - an introduction to structural analysis within sociology, and specifically network theory. I have some problems with a pure structural approach, and the network theoretical view is one with which I feel much happier. He presents six "analytic principles":
A good introduction to the benefits (and pitfalls) of network analysis within sociology. (Wellman, Barry, "Structural analysis: from method to metaphor to theory and substance", in Wellman, B., and Berkowitz, S.D. (1988) Social Structures: A Network Approach, Cambridge University Press, Cambridge, pp. 19-61)
- Ties are usually asymmetrically reciprocal, differing in content and intensity.
- Ties link network members indirectly as well as directly. Hence, they must be defined within the context of larger network structures.
- The structuring of social ties creates nonrandom networks, hence clusters, boundaries, and cross-linkages.
- Cross-linkages connect clusters as well as individuals
- Assymetric ties and complex networks differentially distribute scarce resources.
- Networks structure collaborative and competitive activities to secure scarce resources.
- Etienne Wenger - Communities of Practice: Learning, meaning and identity A brilliant examination of the concept of communities of practice, how they arise, the requirements for their health, and how they encourage learning. It only falls down by trying to make concrete suggestions for implementation/design in what is really a theoretical exploration. There are, possibly, too many concepts thrown in, but the work is excellent, and includes some very concise descriptions of other relevant theories in the endnotes (in particular pp. 279-283). The main concepts that Wenger identifies as necessary for communities of practice are participation and reification (of, for instance, boundary objects). Negotiability of meaning is a core concept, and he believes that without engagement, imagination and alignment by individuals, communities of practice will not be robust. His accent on the importance for identity-formation (creation and growth) of the nexus of communities of practice in which each of us is engaged is, for me, key. A fascinating and perceptive book. (Wenger, Etienne (1998) Communities of Practice: Learning, meaning and identity, Cambridge University Press, Cambridge)
- H. Wichman - Effects of isolation and communication on cooperation in a two person game - Wichman runs a series of Prisonner's Dilemma games with different kinds of communication allowed to participants. He finds that "isolated" players (no visual or auditory contact) have the lowest payback, followed by "seeing only" (visual contact, no mouthing, no writing), then "hearing only", then "seeing and hearing". In my opinion, this leads to the question of what research has been done on how online communication technologies compare to these situations - an area which the "knowledge management" practitioners may have examined. (Wichman, H. (1970) "Effects of isolation and communication on cooperation in a two person game" Journal of Personality and Social Psychology 16, 114-20)
- Alfarez Abdul-Rahman and Stephen Hailes - A Distributed Trust Model Abdul-Rahman and Hailes construct a model for allowing the creation of trust in a distributed system, with recommendations being passed between entities. It's very close to my area of interest, and addresses many areas which concern me. They suggest that entities should hold direct trust values for others, and also recommender trust values (how much do you trust someone to recommend another?). Issues that occurred to me were: how do you define contexts, and how can you correlate another entity's trust aresa with yours (and might they not lie?). There's also the question of how important a particular context is to another entity (and their level of expertise): would you trust a high recommendation about scuba training from a beginner, or from an expert, and how do you express this? Some good points about culture and trust, and an acknowledgement of the importance of trust as a way of managing risk. Lots of references, too. (Abdul-Rahman, A. & Hailes, S. (1997) A Distributed Trust Model, online at <URL: http://www.cs.ucl.ac.uk/staff/F.AbdulRahman/docs/hicss33.pdf>, accessed 2003-04-13)
- Patrick Bateson - Biological Evolution of Cooperation and Trust. Explanation of how cooperatoin and trust in animals is consonant with the theory of evolution, with an description of three possible explanations (all of which he thinks may have played a part): social cooperation as an outcome of competitive evolution (a common example being "kin selection"), the changing of group behaviour in such a way that it is more in an individual's interests to play along than to "defect", and "mutualism within species", where individuals cooperate with each other if they are more likely to survive and reproduce themselves if they help the other. (Bateson, Patrick (1988) "Biological Evolution of Cooperation and Trust" in Gambetta, Diego (ed.) Trust: making and breaking cooperative relations, Basil Blackwell, Oxford, pp. 14-30)
- Partha Dasgupta - Trust as a Commodity. Mainly an (fairly formal) examination of strategies for establishing reputation, which I felt let itself down a little by being too restrictive in the set of circumstances examined, though this was due in part to issues of space and complexity, as mentioned in the text. His seven starting points, however, certainly bear reference:
(Dasgupta, Partha (1988) "Trust as a Commodity" in Gambetta, Diego (ed.) Trust: making and breaking cooperative relations, Basil Blackwell, Oxford, pp. 49-72)
- without punishment, there will not be incentive to fulfill a contract
- the threat of punishment must be credible
- trust among persons and agencies is interconnected (e.g. you must trust the enforcement agency to be consistent)
- you must trust a person to do something not only on his/her word, but because of his/her choice to do it
- you need to look at the world from the other person's perspective (which is why backwards induction is preferably to forwards induction)
- trust's worth can be measured in particular circumstances (I feel that the issue of the specific worth of trust, and how this can fluctuate, is an imporant issue which demands further thought)
- trust only makes sense as a term if it involves expectation of actions before the monitoring of those actions.
- Prashant Dewan - Reputation Systems: Injecting Trust in Peer-to-Peer Systems. An abstract of some work by Dewan on reputation systems as part of PRIDE, referenced at http://www.public.asu.edu/~dewan/Research.htm. Key parts to the system seem to involve trust, identity, accountability and non-repudiation. (Dewan, Prashant (undated) Reputation Systems: Injecting Trust in Peer-to-Peer Systems, online at <URL: http://www.public.asu.edu/~dewan/reputations.htm>, accessed 2003-09-27)
- Magnus Enquist and Olof Leimar - The evolution of cooperation in mobile organisms. Enquist and Leimar posit that cooperation will be difficult to encourage (or maintain as an ESS) in Prisonner's Dilemma-type systems where mobility is high, as free-riders have more chance of defecting and getting away with it. They show that suspiciousness (waiting before commitment to a new partner) and gossiping (members of the population discussing other members) are good ways of helping cooperation. Suspiciousness is a way of demanding a token of commitment (time is very important in the sorts of system Enquist and Leimar are evaluating), and gossip, of course, a way of creating reputation. (Engquist, M., and Leimar, O. (1993) "The evolution of cooperation in mobile organisms" Animal Behaviour 45, pp. 747-57)
- Diego Gambetta - Can We Trust Trust? - a conclusion to the book, which discusses a number of issues around trust. He offers the following (somewhat laboured) definition of trust:
He also suggests that "[w]hile it is never that difficult to find evidence of untrustworthy behaviour, it is virtually impossible to prive its positive mirror image (and quotes the case of Iago's "proof" to Othello of Desdemona's infidelity. (Gambetta, Diego (1988) "Can We Trust Trust?" in Gambetta, Diego (ed.) Trust: making and breaking cooperative relations, Basil Blackwell, Oxford, pp. 213-238)
- "trust (or, symmetrically, distrust) is a particular level of the subjective probability with which an agent assesses that another agent or group of agents will perform a particular action, both before he can monitor such action (or independently of his capacity ever to be able to monitor it) and in a context in which it affects his own action."
- Diego Gambetta - Mafia: the Price of Mistrust - an analysis of the mafia in southern Italy and Sicily, and how the mafiosi thrive on an absence of trust. He suggests that the mafia works through 1) fear of sanctions, 2) because cooperation enhances people's mutual economic interests, 3) because people have general reasons (culture, moral or religious) for believing that cooperation is good irrespective of sanctions and rewards, and because 4) they are related by bonds of kin or friendship. That the mafia inject mistrust into society for their own goals seems certain - however, in a non-governmental system (such as a p2p network), to what extent can we always (or even ever) privilege one type of trust network over another? If one network provides resilience, maybe it is stronger, fitter as a population. Obviously, we need to understand what we want of the system, and how we "rate" trust models. (Gambetta, Diego (1988) "Mafia: the Price of Mistrust" in Gambetta, Diego (ed.) Trust: making and breaking cooperative relations, Basil Blackwell, Oxford, pp. 158-175)
- Ernest Gellner - Trust, Cohesion, and the Social Order. Gellner looks at trust in Middle Eastern Islamic societies which have a strong urban/pastoral dichotomy, and aruges that anarchy engenders trust (or social cohesion). Clans get together to stop stealing of flocks (which is easy) by making it clear that retribution will happen should theft occur, and pledge their (fluid) allegiance to larger tribal groups. His thesis (from Ibn Khaldun) is that urban centres do not engender trust, as clan or kinship bonds are less strong, and debts (such as dowry debts) are more economically based, and of a shorter term, than in a pastoral society. Ruling elites therefore grow weaker in the city until a new, previously pastoral, elite sweeps into town and takes over until it, in turn grows weaker and the cycle repeats. (Gellner, Ernest (1988) "Trust, Cohesion, and the Social Order" in Gambetta, Diego (ed.) Trust: making and breaking cooperative relations, Basil Blackwell, Oxford, pp. 142-157)
- David Good - Individuals, Interpersonal Relations, Trust. An excellent introduction to some of the reasons why establishing, maintaining and demolishing trust are not as easy as might be expected, and are affected strongly by psychological and sociological issues. Major points include: the extent to which theoretical findings are often based on short-term laboratory tests, how removing the focus on "threat" in a set of relationships raises the chances of cooperation, and how small-sum increments in reward are a good incentive to cooperate. He also argues strongly that many decisions, such as those about how to assess reputation, are not made rationally, and that even if our first set of decisions are made rationally, we are very good at convincing ourselves to stick with decisions (or strategies) despite good evidence to the contrary (as suggested by theories of cognitive dissonance, for example) - this is confirmation bias, and bias towards preservation of a theory is often referred to as the "set effect" or Einstellung (he suggests that this underlies the "small-sum increments" behaviour Also important is the context in which decisions to trust are made, and the recognition that these contexts may be different for the different parties (as was the case when Native Americans "sold" land to white settlers, without sharing the concept of ownership of land at all). (Good, David (1988) "Individuals, Interpersonal Relations, Trust" in Gambetta, Diego (ed.) Trust: making and breaking cooperative relations, Basil Blackwell, Oxford, pp. 31-48)
- Keith Hart - Kinship, Contract, and Trust: The Economic Organization of Migrants in an African City of Slum. Although largely concerned with the issue of how a society largely outside the reach (and culture) of a contractual legal system copes with issues of trust (particularly for credit), Hart's discussion of where our word "trust" is situated is of particular interest. "Trust implies depth and assurance of such feeling [belief as a feeling which varies inversely with proof], with inconclusive evidence or proof." (p.189) He situates it between faith and confidence on a "continuum of words for belief mixing extremes of blind faith and open-eyed confidence." (p.189) He also makes a close association between trust and the "notion of friend." (Hart, Keith (1988) "Kinship, Contract, and Trust: The Economic Organization of Migrants in an African City of Slum" in Gambetta, Diego (ed.) Trust: making and breaking cooperative relations, Basil Blackwell, Oxford, pp. 176-193)
- Geoffrey Hawthorn - Three Ironies in Trust. Looking mainly at trust in political systems, Hawthorn suggests that a "socially extensive trust" - more than a trust between friends - can only be created in and by "aristocracies"; that the members of these aristocracies are likely to undermine it; "that if trust is to be maintained, this will have to depend on conditions which are external to the social arrangements in question". Although his (somewhat broad) definition of aristocracies is interesting, it is not very applicable to the p2p systems that interest me. (Hawthorn, Geoffrey (1988) "Three Ironies in Trust" in Gambetta, Diego (ed.) Trust: making and breaking cooperative relations, Basil Blackwell, Oxford, pp. 111-126)
- Ben Houston - Spontaneous Trust Houston lists some thoughts about the spontaneous growth of trust in a distributed network. Central to his thoughts are that new entities should start with zero levels of trust, and should accrue trust over time by conforming to rules of conduct agreed with other entities. A set of sketched ideas, some of which are interesting, but most of which are lacking in form. (Houston, Ben (2000) Spontaneous Trust, online at <URL: http://www.exocortex.org/p2p/spontaneous-trust.html>, accessed 2003-04-13)
- Raph Levien - Advogato's trust metric In this online page, Levien describes the trust metric used on the online community Advogato, in terms of graph and network theory. Without a better understanding of network theory, I found this rather hard going, and I need to reread it in the future. (Levien, Raph, Advogato's trust metric, online at <URL: http://www.advogato.org/trust-metric.html>, accessed 2003-04-13)
- Edward H. Lorenz - Neither Friends nor Strangers: Informal Networks of Subcontracting in French Industry An excellent study of an situation of imperfect knowledge and incomplete rationality where trust arises without the need for friendship or kinship bonds, in manufacturing firms in Lyons in the mid-1980s. (Lorenz, Edward H. (1988) "Three Ironies in Trust" in Gambetta, Diego (ed.) Trust: making and breaking cooperative relations, Basil Blackwell, Oxford, pp. 194-210)
- Steven Nock - The Costs of Privacy: Surveillance and Reputation in America - ISBN 0-202-30454-X (hardback), 0-202-30455-8 (paperback). Nock takes as his thesis the proposition that the number emancipated young males and females (i.e. those living away from home) has risen steadily over the past 100-150 years, and that the erosion of control via "shame" that a family-based society can provide has led to its replacement with a "guilt" system. Such a system requires "strangers" (a growing group) to carry reputations with them so that people can make a decision as to whether or not to trust them. He argues that "credentials" and "ordeals" are two ways of providing the surveillance that such a reputation system requires, and attempts to correlate the rise in a variety of these tools (credit cards, driving licenses and educational degrees on one hand, and lie detector tests, integrity testing and drug testing on the other) with the rise of emancipated young males and females, with some (but not total) success. He goes on to discuss genetic testing and where it might take society in the future, and Bogard's extension of this thread into profiling is worth following. One major concern with Nock's thesis is that he doesn't address other external factors, and doesn't seem to deal with the fact that the correlation between some of his variables is very low. More important for me, however, is that he hardly mentions at all punishment and the enforcement of systems which provide credentials. An interesting read nevertheless. (Nock, Steven (1993) The Costs of Privacy: Surveillance and Reputation in America, A. De Gruyter, New York)
- Anthony Pagden - Trust in Eighteenth-century Naples. An account and analysis of how the ruling Spaniards systematically destroyed "fede pubblica" - public trust (or maybe better "social capital") - by setting up "honour" as more important than justice and the common good. They "destroyed the rule of law by setting up separate courts for both the barony and for the priesthood and by allowing into the legal system entire categories and exemptions and exceptions to that no one could predict the outcome of a case or know which part of the law applied to him." He also believes that setting up women as objects of love, rather than respect, and thereby destroying normal interactions between the sexes, helped break down social trust. (Pagden, A. (1988) "Trust in Eighteenth-century Naples" in Gambetta, Diego (ed.) Trust: making and breaking cooperative relations, Basil Blackwell, Oxford, pp. 127-141)
- Piotr Sztompka - Trust, A Sociological Theory - ISBN 0-521-59858-0. An introduction to the sociological theories of trust. Very much grounded in meatspace, rather than cyberspace, but a useful introduction nevertheless. I find his definition of trust a good starting point for discussion:
(Sztompka, P. (1999) Trust: A Sociological Theory, Cambridge University Press, Cambridge)
- "Acting in uncertain and uncontrollable conditions, we take risks, we gamble, we make bets about the future uncertain, free actions of others. Thus we have arrived at the simple, most general definition of trust: TRUST IS A BET ABOUT THE FUTURE CONTINGENT ACTIONS OF OTHERS." (p. 25)
- Bernard Williams - Formal Structures and Social Reality. Looks at alternatives to the classical Prisonner's Dilemma (PD), such as Sen's Assurance Game (AG), and Sen's Other-Regarding game (OG), and the implications for repeated games in societies where these games predominate. In particular, he examines to variables on motivation: "macro/micro" (whether motivation is fixed on particular game or on a set of games), and "egoistic/non-egoistic" (whether motivation is non-altruistic or altruistic/ethical. For instance, in a Hobbesian structure, we would expect egoistic macro-motivation do predominate. He concludes that none of the four possibilities suits our modern society at the level of generality usually discussed. He looks for "thick trust" situations, between smaller groups of people and an accent on more personal, rather than impersonal interactions. (Williams, Bernard (1988) "Formal Structures and Social Reality" in Gambetta, Diego (ed.) Trust: making and breaking cooperative relations, Basil Blackwell, Oxford, pp. 3-13)
home : about : p2p : security : technology : trust
Mike Bursell - email@example.com
Last modified: Sun Apr 13 17:26:49 BST 2003