The brilliant Master Shirky offers a stunning critique on the limitations of RELATIONSHIP (with clarifications).
Key Shirky views:
– “a formal and explicit ontology for human relations is unworkable”
– “most human relations cannot be made explicit without changing the nature of the relationship”
– terms for classifying relationships are unbounded (and hell, people can’t do do it anyway)
– FOAF developers can’t develop their own ontology because of their insider role
Not only do i agree his views on the matter, i think that they need to be affirmed. I’m also SUPER psyched that he referenced that AI debates because this historical precedent is crucial for understanding why so many of the discussions around social software are flawed.
I would also like to add a few additional points to why this problem is unsolvable:
Relationships are situated within a CONTEXT.
Think about the times when you’ve introduced somebody differently to different people. Here’s an example. Said to boss: “Alex is my friend.” Said to best friend: “Alex is this girl i’m fucking.” Said to mom: “Alex is this nice girl i’m dating.” Which is it? All? None? Context!
Another context: time. Your relationship with someone changes over time. Duh. But guess what? It also changes over local periods of time. For example, i can label the person in my kitchen right now as my ex. I can also label him as tonite’s chef. When i’m done with this entry, i’ll probably label him my confidante. His role is not consistent.
Relationships are defined by CULTURE; their types are SOCIAL CONSTRUCTS.
The term “friend” means different things in different cultures. Hell, even the term cousin differs. In fact, if you want to have a field day, check out anthropology kinship research. For some cultures, what we might call “uncle” another culture would call “father” (an individual would have multiple fathers). To define a universal relationship structure is to project our cultural norms onto other peoples. Yet, without a universal structure, there’s no common language.
On a more personal level, how many of you have ever called your step-mum mum? Are you lying? How would you categorize her here? What about your adoptive mum? What are the implications for you, for her, for others? People label their step-mums and adoptive mums as mum as a social construct to indicate the value of the relationship, not simply a reflection of the biological term. Likewise, there’s nothing worse than telling someone “you’re just my step-mum” or “you’re not my real mother; you just adopted me.” What are the implications for using those terms in an ontology?
Relationships do not exist without POWER.
No relationship exists without power (see Foucault’s “History of Sexuality v1”). Power can be shared via turn-taking, but there is no such thing as pure equality in a relationship. There are times when one person has power over another; sometimes, the reverse is true. Relationships are a negotiated process.
While labeling some relationships, the power dynamic is unveiled. For example, i cannot be the mother of the person who is the mother of me. Inherently, there’s a mother/daughter relationship, marked explicitly by its power. Some such relationships are only 1-way: fan/stranger. The obsessed doesn’t know the obsessor.
Then there are the relationships that we’re determined to define equally. Friend-Friend. Do both people get the same thing out of the relationship? Does “Friend” mean the same thing to each person? Hell, think of all of your friends. Do you really mean the same thing when you label dozens of different people with that term? (If you do, please seek therapy.) We throw that word around because often the process of making explicit the power dynamic is neither socially acceptable nor something we want to own up to.
Relationships are often built on an undiscussed meeting of each other’s needs. “She’s my friend because she always gives me a ride home. In return, i listen to her bitching about her job.” Both people are getting something out of the relationship that they each need/want. It’s valuable to each of them, but neither really wants to make that explicit.
The best debate on this inevitably surrounds sex. Sex workers have the negotiation down. Money for sex. People may scoff at this explicit negotiation, but many of us have had sex for far less honorable reasons. Ever had sex with your lover so that s/he’ll go to sleep? Ever had sex to spite someone? To get what you want? Sexual relationships emerge from power dynamics. Rarely do people engage in sex for the exact same reason. To get off? To feel loved? To feel validated? The BDSM community recognizes this power dynamic and makes it explicit; most of us do not.
Given that most of us aren’t really able to address our power issues, how are we supposed to label them?
Update: Ideas Bazaar discusses this in terms of kinship terminology, anthro style. The focus is on lack of quality terminology. [via Foe]
danah,
a quick add to your discussion of power above. There’s always a power dynamic (asymmetry) in the relationship, yes, but even more to the point is the question of who’s allowed to name the relationship. I think of my experience in grad school, and now my relationships to the people I mentor (as a prof in a doctoral program). I can describe them (i.e., I have the power/privilege of describing them) as students, friends, colleagues, “apprencticesOf,” etc., but they’re in the same position I was–I’d feel odd about introducing certain members of my dissertation committee to others as “my friends,” even in cases where there is friendship there (and the teacher-student dynamic several years past, now).
Power requires us to ask who gets to use the vocabulary as well as whose use of it will receive the most attention, respect, validation, etc…and that swings back into context & culture…
cgb
The contexts in which relationships are situated seem to be constructed, mediated, and even executed as language.How can it be otherwise?
Unless Alex is actually being fucked “as event” in front of the best friend then it is a
consciouly constucted representational narrative being delivered to the best friend (with, as you say, contextual considerations).
One that is of no more, could one say truth value than the description provided to the mother or the boss; all are equally valid. When this representation is removed from context then the nature of the relationship becomes apparent through a form of negation or default.
I thought of the example of the British ‘lad’ who recently (Dec. 2003) distibuted some emails from his oral aquaintance/girlfiend/worl collegue in what the BBC called a “salacious incident” (again a cultural context of langauge I believe..). He distibuted them to some of his mates, they distributed it further and so on…over 1 million mails. The mails became a public humourous confessional semi-porn piece and the represented relationship changed as a result.
The relationship changed through language to reflect the nature of the network. Or was it always a semi-public sexual exchange but just never acknowledged in langauge? Was it the nature of the language in the context (mails originally sent at work) that forced the change in the relationship? Or the medium employed to distribute them?
“She is horrified by all this. She has gone into hiding. I don’t blame her. Would you stick around after something like that?”
Clare Swire’s mother
The mails: http://whoisclaireswire.terrashare.com/
In response to Collin: the power structures in academia here in Sweden are much less formal than I have encountered in other parts of the world (England, Australia and India). It seems to make for a much freer (Friere)language enviroment:)
I like a lot of Clay Shirky’s critique and very much agree with your points, but I disagree with Clay’s starting point–the assumption that a machine readble vocabulary for classifying human relationships is altogether some new frontier (destined for failure).
There are millions of databases–some running for many decades that each:
-use a formal and explicit ontology for human relations
-make human relations explicit without changing the nature of the relationship
-use bounded terms for classifying relationships
Vocabularies are themselves situated within a context. Machine vocabularies as well.
I would argue that it is the removing of context for the vocabulary (or, really, the suggesting that a vocabulary could be unbounded by any context, or it could be universal) that is the “wrong path” that leads to the failures you and Clay Shirky indicate.
The Internet / web, in the way it is universal, isn’t bounded enough to function as a context in which a vocabulary is effective (e.g., consistent, comprehensive, etc.). So I sympathize with reactions to “universalizing” vocabularies that might be very useful in a particular context, but might seem wrong or weird outside of that context.
I like a lot of Clay Shirky’s critique and very much agree with your points, but I disagree with Clay’s starting point–the assumption that a machine readble vocabulary for classifying human relationships is altogether some new frontier (destined for failure).
There are millions of databases–some running for many decades that each:
-use a formal and explicit ontology for human relations
-make human relations explicit without changing the nature of the relationship
-use bounded terms for classifying relationships
Vocabularies are themselves situated within a context. Machine vocabularies as well.
I would argue that it is the removing of context for the vocabulary (or, really, the suggesting that a vocabulary could be unbounded by any context, or it could be universal) that is the “wrong path” that leads to the failures you and Clay Shirky indicate.
The Internet / web, in the way it is universal, isn’t bounded enough to function as a context in which a vocabulary is effective (e.g., consistent, comprehensive, etc.). So I sympathize with reactions to “universalizing” vocabularies that might be very useful in a particular context, but might seem wrong or weird outside of that context.
Shirky on the Relationship vocabulary
After seeing and commenting on danah boyd’s RELATIONSHIP: Context, Culture, Power post, I felt like I should retrieve some of my letter, put it a little differently, and publish it here.
Social Context Redux: User versus Network Centric
Clay Shirky and danah boyd both comment on the problems of modeling social constructs. Their discussions highlight a common problem with today’s YASNS: Network-Centric Design. As systems focus on the connections between people, they neglect to accoun…
FOAF developers can’t develop ontologies?
Since I’ll be hanging with both Clay and danah next week at a hoidy toidy social software scene (with Linda Stone no less) – I think I’ll put up a placeholder and say (as a FOAF developer) : “we can develop many kinds of ontologies and let our end-users d
Danah Boyd on RELATIONSHIP
Danah Boyd has posted a thought-provoking criticism of the RELATIONSHIP ontology (by way of Marc Canter) While I approach this matter from a strongly technical perspective, the perspective she presents should be carefully considered as we define social…
Danah, very well put! I am now thinking about this issue in a different way.
Snipped and condensed from my long blog posting:
The last point particularly struck me. The act of describing a relationship, much less having a relationship, is a transaction of power.
My answer is that we will have to allow the semantics of relationships to emerge from the somewhat distorted political dynamics of the labeling process.
With a more expressive ontology for defining relationships (certainly more so than RELATIONSHIP), we may actually end up with a better understanding of the social dynamics of the real world, across the world. In order to have a conversation, we need to agree on some conventions, so we can’t complain too much about the RELATIONSHIP. 🙂
clear definitions of relationships are hopelessly conflated with language, and the ambiguity of language automatically implies problems. many people seem to be vigorously agreeing on this point.
this (semantic) language issue could be taken, by extension, to include culture; you express this in your second point above – that relationships are tied to culture and are a social construct. while this seems valid, i wonder whether there are some ethnographers who might argue.
specifically, there does seem to be a body of evidence that a set of human universals exists, independent of culture and language (e.g., D.E. Brown, Human Universals, 1991; or Pinker’s discussion of nature, nurture, and human universals in The Blank Slate). brown’s list consists of primarily surface universals of behavior and overt language. nonetheless, many of these universals are clearly relationship-oriented.
mr. pinker might also argue with you about whether or not relationships are entirely social constructs, but i’d be best to leave that discussion to the two of you, should you ever meet. 🙂
::laugh:: I have met Mr. Pinker, although i could not describe our relationship. He has no idea who i am and i have mixed opinions of him as a teacher vs. writer.
But yes, there’s a reason why a rift emerged between cognitive science and psychology. Us social scientists think that most things are socially constructed.
Kinship, Logic and Games
All this kinship and networks talk of late is making me realise that I should have been listening at University during kinship courses. Shame on me. It has also forced me to dig out old papers I wrote, from which…
Kinship, Logic and Games
All this kinship and networks talk of late is making me realise that I should have been listening at University during kinship courses. Shame on me. It has also forced me to dig out old papers I wrote, from which…
Kinship, Logic and Games
All this kinship and networks talk of late is making me realise that I should have been listening at University during kinship courses. Shame on me. It has also forced me to dig out old papers I wrote, from which…
Kinship, Logic and Games
All this kinship and networks talk of late is making me realise that I should have been listening at University during kinship courses. Shame on me. It has also forced me to dig out old papers I wrote, from which…
Kinship, Logic and Games
All this kinship and networks talk of late is making me realise that I should have been listening at University during kinship courses. Shame on me. It has also forced me to dig out old papers I wrote, from which…
Kinship, Logic and Games
All this kinship and networks talk of late is making me realise that I should have been listening at University during kinship courses. Shame on me. It has also forced me to dig out old papers I wrote, from which…
socially constructed
Oh dear. Now if you’d said conditioned by their economic substructure, then maybe I could go along with you.
Relatedly yours
Reflecting on the classic relationship vocabulary flap of just last week, I am shocked, shocked! that no one (that I
Relatedly yours
Reflecting on the classic relationship vocabulary flap of just last week, I am shocked, shocked! that no one (that I
Relatedly yours
Reflecting on the classic relationship vocabulary flap of just last week, I am shocked, shocked! that no one (that I
Who do you know?
Roundup of some recent discussions about codifying relationships in online social networks
WBC04: day 2 morning
Developing bulletin board visualizations (.pdf)
A lot of this discussion seems to hinge on what one means by logic, and the relationship between form, interpretation and use.
The terms comprising a kinship terminology can be described by a logic constructed without reference to genealogy using user judgements on products of the terms themselves.
However, this logic only defines the terms and the products of terms (e.g. father of father = grandfather, which is derived from x calls y father, y calls z father, x calls z grandfather). This only governs this form, and a ‘default’ manner of instantiating this logic over a given group of people and their relatives. The logic is different for different terminologies, though there are a relatively few ‘families’ of logics. So far in our research all of these logics have additionally comformed to structural criteria which defines them as algebras. The semantics of such a terminological logic (as with any logic), is simply the assignment of a product to a value (kinterm/not a kin term in this case). Meaning is relegated to other processes.
This does not limit in any way how these terms are instantiated (since the terminological logic is determined by the internal relations between terms, not against geneaological data). So even people who might use ‘mum’ for ‘step-mum’, will make the same judgements of relations between terms divorced from specific people. The instantiation of mum for step-mum is determined by non-terminological issues.
What it does do is define a basic structure for people to build on that can be distributed with high fidelity. Even if a terminological logic is unique to a people, to use it effectively a solid foundation has advantages over people having to memorize for a group of 1000 people they live with 999,000 different relationships, instead only needing to learn between 2 and 20 terms, or so, which combined with the specific logic of their terminology allows people to determine a relationship between themselves with a single common relationship (or absense thereof).
Instantiation (sometimes, not quite correctly, called pragmatics) relates to how people actually use these relationships. Parts of the process of instantiation are probably driven by at least strategies, if not logics, and represents a lot of different (and not always consistent) cultural systems interacting together, some more arbitrary and personal, others shared to a considerable extent.
What does this have to do with the prospects for ‘relationship’ vocabularies? I agree with Shirky that the Relationship vocabulary is a loser. It is a list of terms, with no particular structure between them, allowing no baseline interpretation upon which to build the specific meanings that arise in instantiation or use that form most of the comment on this page. However, it would be possible to build such a beast as a combination of logic and terminologies. However, it might not be possible to build a universal one, as the relationships between these terms will vary from culture to culture (and more and less terms will be necessary). This is a general problem for all relational schemas, and current XML related technologies are not up to the job in their present formulations, in particular RDF.
However, it is probably possible to build a formal description of each one, and to represent this in a form that permits comparison, and in many cases translation. This requires a more dynamic form of Resource description than currently exists. This is a thread we are working on at the included URL. But a single set of definitions for any relational terminology is fruitless unless there is a single logic underlying these.
Relatedly yours
otherwise platform-independent, be the.