A deliberative structure or deliberative framework or deliberation framework or argumentation framework is central to an online deliberation scheme. It allows participants to express their views in a semi-structured way (that is, a little more disciplined than plain text and a little less disciplined than formal databases).
The best-known lineage of deliberative structures was started by Horst Rittel at Berkeley in the US in the 1970s; It was eventually developed into the gIBIS hypertext system in the 1980s, a predecessor to the World Wide Web that supported many semantic web features. The most interesting of which was the division of statements into issues and positions so that a neutral issue statement and biased position statements could be accomodated on the same page. There were two direct descendants of this work by 2007:
- The question/idea/pro/con model championed by Jeff Conklin, suitable for dialogue mapping and early brainstorming, which he had developed through his work at CM/1 and MCC, now expressed in his proprietary Compendium system
- The issue/position/argument or issue/policy/argument model implemented by the Green Party of Canada Living Platform, Living Agenda, openpolitics.ca, and let.sysops.be, first documented fully (as time/issue/position/argument/source/evidence/authority form by anonymous trolls at meta.wikipedia.org; This mimics the classic deliberative structure used in adversarial process such as a criminal court, and merely accelerates it via online deliberation in mediawiki or tikiwiki, for which software the conventions have been must fully developed to date.
 Arguments for IPA
A detailed comparison of the IPA, debatepoint.com and other forms favours the use of IPA for civic work for these reasons:
- Issues are presumed to be stable and persistent, and to change only slowly with broad public opinion or scientific consensus, possibly over decades - this favours input from a huge number of people over a long period of time and makes deep framing by one faction more difficult than in the other forms which favour moderators and regular users
- Positions are not just "ideas" (possibly originated by crapflood or astroturfing) but are ideally actual committed political or legal positions to which someone has attached a name or reputation - for instance, the positions taken by parties during actual elections
- When the legitimacy of an argument is questioned, IPA allows for regress along the same lines exactly as academic matters require: evidence/source/authority. First evidence is examined, and also its source if that is questioned, and ultimately an authority that can state the source to be valid or unbiased (e.g. a magistrate) may have to declare whether a source is valid or not. For instance, whether evidence came via legal channels or not.
- IPA maps easily onto meeting agendas as was done by openpolitics.ca and facilitates pre-existing organization protocols as required by corporate or party or agency charters.
There are also reasons to believe that using IPA will make it easier to align wikis:
- IPA was developed by people pursuing broad and basic civic causes to which they were and are committed and continues to be developed by such people, not by dilettantes pursuing academic, corporate or other agendas narrower than those of the developers of IPA/TIPAESA
- Much material in IPA form describes political issues. The featured issues list of openpolitics.ca or position namespace of dkosopedia.com for instance include some of interest also in the UK.
- Almost all of the most useful material on IPA is under Creative Commons licensing.