A Social Environment Model of Socio-technical Performance

Brian Whitworth

Institute of Information and Mathematical Sciences, Massey University, Albany, New Zealand

Socio-technical levelsLevel Discipline System Combination Examples

Community Sociology, Politics, Business

Social Socio-technical Systems (STS)

Culture, roles, laws sanctions

Individual Psychology, Biology

Cognitive Human Computer Interaction (HCI)

Attitudes, beliefs, ideas, opinions

Informational Computer Science, Information Science

Software (S/W) Technology

(H/W & S/W)

Programs, data, bandwidth, memory

Physical Engineering, Physics, Chemistry

Physical

Hardware (H/W)

Computer, mouse, wires, printer, keyboard

Table 1. Socio-technical levels

Homo-Economicus

• Individual does what benefits themselves, by reduced effort, increased gain, or both

• Mill’s economic man, who seeks wealth, leisure, luxury and procreation

• Competition for limited resources creates a need for competence

Individuals Competing Model

Figure 1. Individuals competing in a world environment

Rule 1: The Selfish Rule• Freely acting individuals {I1, I2 …} face action

choices {a1, a2 …} with expected individual unit values outcomes {IU(a1), IU(a2), …} follow the rule:

If IU(ai) > IU(aj) an individual should prefer ai over aj Selfish individual choose acts expected to give more value to yourself.

• A defeasible rule • Value includes psychological gains like appreciation,

or social gains like reputation

Homo sociologicus

• Our bodies are cooperative cell colonies, with cancer what happens when cells “defect”

• Ants and bees form massively cooperative societies by genetics - the competing evolutionary unit is the colony not the individual, i.e. Biologists now argue for multi-level selection

• Marx’s communist man

• Social cooperation creates synergy benefits

Communities Cooperating Model

Rule 2: The Social Rule• If a social unit S of { I1, I2 …} individuals faces

social action choices {a1, a2 …} with expected social unit values of {SU(a1), SU(a2), …}, then:

If SU(ai) > SU(aj) then prefer ai over aj

• Socialized individuals choose social acts expected to give more value to the community

Note: Social acts reference social units not individuals, e.g. “defend society” is independent of individual state. Allows social “castes” like worker or soldier

Synergy

• Difference between what individuals produce as a social unit vs what they produce alone

• Trade illustrates positive synergy

• Conflict illustrates negative synergy

• Generally pays individuals to join positive synergy social units, and leave negative ones (they are better off alone)

• A property of the number of interactions, not the number of group members

The Social Dilemma

• While genetics drives ant society, people can choose to follow Rule 1 or 2

• What if Rule 1 conflicts with Rule 2?i.e. what is good for society is not what is good for the individual?

• Only Rule 2 allows synergy gains, but Rule 1 is the primal rule in nature

The Prisoner’s Dilemma

• Prisoners Bill and Bob face two year jail for a crime they did commit

• Each can plea bargain to testify against the other

• If Bill testifies and Bob doesn’t, he walks free and Bob gets 7 years jail

• If both testify, both get six years (one off for testifying).

Bob Years free (Bill/Bob) Cooperate Defect

Cooperate 5/5 0/7 Bill

Defect 7/0 1/1

By Rule 1 it always pays individuals to defect

Other Social Dilemmas

• Tragedy of the commons: – Farmers by a commons with cows and a land plot– If a farmer grazes the commons, his herd grows fat– If all farmers do so, it is overgrazed and dies off– Parallels modern conservation problems

• Volunteer dilemma• Social loafing• False representation, etc, …• Individuals alone can’t solve social dilemmas,

one “do gooder” is just a “sucker”

Social Instability

Valleys of Defection

• Anti-social acts like stealing “short-circuit” synergy gains• Each defection reduces synergy in a cascade effect •Rule 1: Synergy is unstable

Peak of Synergy

Zero-Sum BarrierZero-sum:

Expand your slice – world domination!

Non-Zero-sum:Expand the pie – to expand your slice!

Human civilization somehow achieved massive non-zero-sum gains by non-genetic means

Social Order• In perfect social order all individuals are “one

mind”, cf in a crystal all atoms move as one

• Social anarchy- gas atoms move individually

• A community with social order, by religion, culture or laws, avoids stealing and cheating (social disorder)

• Can solve the social dilemma by following Rule2, but at the expense of freedom/Rule1

• “Barbarians” (Rule 1) vs “Civilization” (Rule 2)

Social inventions • Unfairness. Not inequity—unequal distribution

of outcomes—but not distributing outcomes according to contribution, e.g. that fit adults live idly while others work to support them is unfair

• Justice—punish unfairness so Rule 1 no longer profits—social order plus individual freedom– Social unit transmits world requirements

(accountability)– People have a natural justice perception– Revenge is a primitive form of justice – State justice (police, laws, courts, prisons) aims to

deny unfairness (Rawls, 2001)

Social Hijack

• Individuals take social control for their own ends, just as a virus hijack a cell

• Benevolent dictators (Plato) enforce social order (synergy), then justly return the gains to society

• Dictators keep control by repressing and indoctrinating• Dictatorships are:

– Unstable. Slaves have “nothing to lose but their chains” Marx – Impermanent. Kings, emperors, pharaohs, etc die, leaving a

power vacuum. Bloodline dynasties over time produce incompetent offspring

– Unproductive. In Zimbabwe Mugabe addressed social inequity by driving white farmers off productive farms, then gave them to cronies who looted - turned Zimbabwe from the bread-basket of Africa into the basket-case of Africa.

The golden rules• Do unto others as you would they do unto you• Rabbi Hillel’s sum of all rules: “If you don’t like it

done to you, don’t do it to others”. • Kant’s proposal: “Act only on that maxim by which you

can at the same time will that it become a universal law”, i.e. if everyone does it, is it still successful?

• Pareto’s optimality principle: “Good actions benefit at least one other and do no harm.”

• Rawl’s “veil of ignorance” requires state justice to be “blind” to individual needs.

• Harsanyi rules out immoral or anti-social acts (Harsanyi, 1988).

All encourage free individuals to choose Rule 2

Social environment Model

I1

COMMUNITY ENVIRONMENT

SocialActions

Social Tokens

Cooperation &

Competition

WORLD ENVIRONMENT

Social & Competence Requirements

Social Outcomes

PerformanceRequirements

CompetitionCompetence + Synergy

INDIVIDUAL TECHNICALSUPPORT

SOCIO-TECHNICALSUPPORT

Anti-social acts

Social Barriers & Sanctions

Cheating value

I1 ...

...

...

...

Figure 3. Social environment model

Good Citizen Rule 3Rule 3

a. If {SU(ai) ≥ SU(aj) and IU(ai) > IU(aj)} then prefer ai to aj

Choose acts that don’t harm society significantly, but benefit oneself

ORb. If {IU(ai) ≥ IU(aj) and SU(ai) > SU(aj) } then prefer ai to

aj Choose acts that don’t harm oneself significantly, but benefit

society

Rule 3 is a hybrid of Rule 1 and 2

Self vs Community Choices

COMMUNITY

Gain Minor effect Loss

Gain Synergy Opportunity Anti-social

Minor effect Service Null Malice

S E L F Loss Sacrifice Self-harm Conflict

Rule 3 favors service, synergy and opportunity

Socio-technology

• Online people help others they have not met and may not meet again, Neither Rule 1 nor Rule 3a explain this

• Socio-technical systems succeed by good citizens – “small heroes” doing small selfless acts for others

• That virtue” is productive and supportable by technology is an important social discovery (Benkler & Nissenbaum, 2006)

• Socio-technical systems are a new social form, that change the social focus from denying defection to enabling good citizenship, e.g. open source

Examples

Aim Examples Synergy Defection

Communicate Email, Chat, ListServ, IM

Shared communication: People send messages they otherwise would not

Spam: Spammers waste others time, giving spam filters.

Learn WebCT Moodle Blackboard,

Shared learning: Students help others learn, reduce teacher bottlenecks

Plagiarism: Students copy other student’s work, giving systems like Turnitin.com.

Knowledge Wikipedia, Tiddlywiki

Shared knowledge: Taps knowledge of the group, not just a few ”experts”

Trolls: Wikipedia’s monitors and rights fight “trolls” who damage knowledge.

Friends Facebook, Myspace

Relationships: People keep in touch with friends and family

Predation: Social network predators find victims, giving reporting and banishing

Keeping current

Digg, Del.icio.us

Shared bookmarks: Social bookmarks let people see what others look at.

Advocates: Who “digg” a site because of a vested interest, e.g. they own it.

Play Second Life, MMORPG, Sims

Shared play: An avatar experiences things impossible in reality.

Bullies/Thieves: “Newbies” robbed by veterans don’t return, so need “safe” areas.

Trade E-Bay, Craig’s List, Amazon

Item trading: People from anywhere exchange more goods.

Scams: Scammers are reduced by online reputation systems.

Work Monster Work trading: People find and offer work more easily.

Faking: Padded CVs and fake job offers need online reputation systems.

Down-load Webdonkey, Bit-Torrent Napster,

Shared down-loading: Groups share the processing load of file downloads.

Piracy: Napster was in conflict with society’s copyright laws, so closed down.

Media Sharing

Flickr, YouTube podcasting

Shared experiences: People share photos/videos with family/ friends.

Offensiveness: Editors remove offensive items—violence, porn, scatology…

Advice Tech help boards like, AnandTech

Shared technical advice: People who have solved problems can help others more easily.

Confusers: People who start new tracks rather than checking existing ones are relocated and scolded.

Express opinions

Slashdot, Boing-Boing, Blogs

Shared opinions: People express and read others opinions more easily

Caviling: People who “peck” new ideas to death—karma systems deselect them.

Other applications

• Enron – A higher form of cheating• Credit crunch – A higher form of incompetence (in

risk management)• Social inflation – When social environments ignore

the demands of their environment, and social tokens lose external value, e.g. money (social token) loses value relative to external standard of a loaf of bread

• Rectification –the demands of outer environments ultimately “cascade” over inner ones

Modelling Social Behavior??

• Foxes/Rabbits– Move– Predate– Breed

• Social Foxes/Rabbits– As before PLUS– Combine: If both agree, form

combined unit with double rewards plus synergy

– Defect: If in combined state, • Defector gets plus synergy

• Sucker gets minus synergy

In Rule 2 state each creates others synergy