Contribution
The paper names and specifies Issue-Based Information Systems: an information system organised around the issues a planning group argues about, rather than around a documentation thesaurus. It gives the elements (topics, issues, questions of fact, positions, arguments, model problems), the relations between issues, six subsystems, and the main operating cycle. It reports three manually run systems in experimental use. It is a specification with a field report of one sentence, not an evaluation.
Key insight
The unit of a planning record is the open question people disagree on, not the topic or the document. “Issues are the organizational ‘atoms’ of IBIS-type systems” (¶5). A discourse is a counterplay of questioning and arguing in which “every formulation of the problem is also a statement about a potential solution” (¶2), so the record has to follow the argument rather than a fixed classification of its subject.
Assumptions
- Cooperatives with a problem complex. The users are “governmental or administrative agencies or committees, planning groups, etc.” that must arrive at a plan for decision (¶1).
- An argumentative model of problem solving. Problem solving by such groups is taken to be an argumentative process (¶1); the paper asserts this, citing the authors’ own reports, and does not argue it.
- Staffed operation. The operating cycle (¶11) has issues “edited (in contact with Mₖ)” and files “processed by S-61”, which implies staff who maintain the system for the participants; the paper does not say who they are.
Key results
There are no results in the sense of findings. What the paper states:
- The process (¶2). A topic is named by a “trigger phrase” (“Urban Renewal in Baltimore”, “The War”, “Tax Reform”). “Issues are brought up and disputed because different positions are assumed. Arguments are constructed in defense or against the different positions until the issue is settled by convincing the opponents or decided by a formal decision procedure.” Questions of fact go to experts or documentation; “Answers obtained can be questioned and turned into issues.”
- Why not a thesaurus (¶4).
- An issue such as “Shall we legalize private manpower banks?” is not represented by the descriptors “Legalization”, “Manpower Banks”, “Private”.
- “The similarity of an issue with issues dealt with previously and the information used in their treatment” matters more than the documentalist’s description of the subject.
- “The discourse developing is notoriously centering around concepts expressed by ad hoc vocabulary which is rapidly changing, defying thesaurization. Many central terms used are proper names for long stories specific of the particular situation, with their meaning depending very sensitively on the context in which they are used.”
- A further aim is “a more scrutinized style of reasoning which more explicitly reveals the arguments”, and “fast and reliable information on the state of discourse at any time”.
- Properties of issues (¶5). They have the form of questions; they originate in controversial statements; they are specific to situations; they are “raised, argued, settled, ‘dodged’, or substituted”.
- Relations between issues (¶6).
- I₂ is a direct successor of I₁: “I₂ challenges a statement made in support of one of the positions maintained in view of I₁.”
- I₂ is a generalization of I₁.
- I₂ is a relevant analogy to I₁: its arguments transfer “mutatis mutandis”.
- Positions on different issues can be compatible, consistent or incompatible, by the same or another proponent.
- Kinds of issue (¶7). Factual (“Is X the case?”), deontic (“Shall X become the case?”), explanatory (“Is X the reason for Y?”) and instrumental (“Is X the appropriate means to accomplish Y in this situation?”). Each may have a closed set or an open list of positions.
- Other elements (¶8). Topics are “a crude organization principle”. Questions of fact are not assumed controversial; “Doubting the credibility of an answer leads to an issue.” Model problems (location models, cost–benefit models) are general, and their usefulness “depends on the importance of those factors which have not been included”.
- Subsystems (¶9). S-1 issue bank (living, settled or abandoned, latent); S-2 evidence bank (questions of fact, answered and open); S-3 “Handbook” of model problems; S-4 topic list; S-5 issue map, a “graphic display of the state of argument”; S-6 documentation system.
- Numbering (¶10). “All the items are numbered consecutively to ease reference.” Relations are kept as matrices (an I×I matrix per relation, an I×T matrix of issues to topics) used “for tracing connections” and for drawing issue maps; literature is linked by accession number, so “several layers of search networks are established”.
- Operation (¶11). O-1 a participant raises issue I_r under topic Tⱼ on an “issue form”; O-2 it is edited and related to other issues; O-3 an “argument sheet” for each position; O-4 literature search; O-5 the issue map updated; O-6 discussion, ending in settlement, a request for evidence, a challenge to a supporting statement (“GOTO S-1”), or denial of the issue’s significance with substitute issues.
- Field report (¶12). IBIS-1 for a supranational agency’s information policy; IBIS-2 for an interdepartmental committee on a national plan for information networks; IBIS-3 for university planning. “The most astonishing result is the eager acceptance by the users although the implementation induced major changes in organization and working styles.” All are manual; computerisation is being programmed.
Claims
| id | claim | strength | support |
|---|---|---|---|
| C1 | Problem understanding cannot be separated from solution, since every formulation of a problem states a potential solution | weak | asserted, ¶2 |
| C2 | Planning discourse runs on rapidly changing ad hoc vocabulary that defeats a thesaurus | weak | asserted from the authors’ experience, ¶4 |
| C3 | Users of three experimental IBIS systems accepted them eagerly despite the changes in working style they forced | weak | a single sentence of report, no data, ¶12 |
Concepts
- Issue. A question, arising from a controversial statement, specific to a situation, on which positions differ.
- Position. An answer taken on an issue.
- Argument. A statement for or against a position.
- Question of fact (F-question). A question not assumed controversial; doubting its answer turns it into an issue.
- Model problem. A general, situation-free model (location, cost–benefit) defined over a closed set of variables.
- Direct successor. An issue that challenges a statement made in support of a position on an earlier issue.
- Latent issue. An issue filed but not yet raised.
Connections
The paper cites only the authors’ own reports (Kunz and Rittel 1968, 1969, 1970) and Rittel (1966). gIBIS (LIT-762) is its best-known computer implementation. The later design-rationale literature filed under ADR-030 (Grudin, LIT-759, among others) is largely an account of why systems of this kind were abandoned, which this paper, written at the start, does not anticipate.
Bearing on the record
- ADR-032’s scheme against IBIS. The record says “an objection is a claim”. In IBIS a challenge to a statement supporting a position is a new issue, a direct successor (¶6), so objections are questions there, not claims. ADR-030’s summary of the planned schemes as “questions, claims and objections, in the manner of IBIS” is loose on this point: IBIS has issues, positions and arguments, and no separate objection element. IBIS also has what the record lacks: questions of fact kept apart from issues, model problems, and the latent and dodged states of an issue. The record has what IBIS lacks: a credence on each claim, and retirement by status rather than an issue’s being “substituted”.
- CLAIM-180 and CLAIM-182 have a 1970 precedent. IBIS numbers every item “to ease reference” (¶10), and it does so because the discourse’s terms are “rapidly changing, defying thesaurization” and are “proper names for long stories” whose meaning depends on context (¶4). That is the drift CLAIM-180 describes, met with addresses as CLAIM-182 proposes. The paper also wants retrieval by “similarity of an issue with issues dealt with previously” (¶4), so it pairs codes with retrieval by content rather than replacing one with the other. It argues none of this; it states it as design.
- CLAIM-179 and the cost of capture. The founding paper reports adoption, not abandonment: “eager acceptance by the users”, with no account of cost beyond “major changes in organization and working styles” (¶12). Its operating cycle implies staff who edit issues “in contact with” the participant who raised them (¶11), so capture in the original design was already done by someone other than the arguer. That bears on CLAIM-179: the disparity Grudin describes may depend on who the filer is, and IBIS’s first design separated filer and arguer, as an agent filer does. It is one sentence of design, not evidence. The question left open in LIT-758’s takeaways, whether the paper says what keeping an IBIS costs, is answered: it does not.
- QUESTION-037. IBIS’s answer to what makes an argument settle is procedural: an issue is settled “by convincing the opponents or decided by a formal decision procedure”, or dodged, or substituted. It offers no account of an argument’s structure stabilising over time, which is what the question asks. Its issue bank and issue map are the closest thing to a project memory, and the paper reports that they were used, not what they changed.
- No THEORY filed. The paper is a design specification. Its three general claims (C1–C3) are asserted, not supported, and nothing it establishes would hold as an account of what is true. C1 is the seed of Rittel and Webber’s wicked problems, which the record should file from that paper if it is read.
- No instruction for machine-learning practice; nothing for the anthology.
Limitations
- Seven typed pages, with no data, no description of the three systems’ users or duration, and no measure of “acceptance”.
- The theory of IBIS and the manuals are announced as “in preparation”; this paper is an outline of them.
- The operating cycle is described for the “treatment of issues” only.
Open questions
- Who edited issues and kept the matrices in IBIS-1 to IBIS-3, and at what cost, is not said; a later report on those systems would answer it.
- Whether “eager acceptance” survived past the experimental phase is the question the later abandonment literature answers for other systems, not this one.