NorMAS-09 was the 4th NorMAS event, the second one at Dagstuhl, where the community working on normative multiagent systems had again the opportunity to meet and discuss.

The seminar was organized around short paper presentation sessions, from 9hrs to 15.30hrs and the remaining time devoted to group discussion. Each paper presentation was commented by a discussant ---who had read the paper and sent comments to the authors beforehand-- and then open to general debate. Priority was given to young and/or female researchers as presenters and discussants.

The papers presented were collected in DROPS proceedings before the meeting (to allows discussant to review them and to make them available to the other participants). A WIKI page was set up for participants to express the issues and technical concerns they find worth discussing in the seminar, and to give access to other seminar-related material.

Afternoon group discussion sessions were structured in two stages. The first one divided the participant into four groups and each group took the current concerns of the NorMAS community --previously gathered in the seminar WIKI-- along with the actual contribution of the papers, and attempted to match concerns and contributions with the challenges of the field that had been proposed in the previous Dagstuhl seminar (NorMAS-07). The second stage consisted of a report of each group and a collective discussion. Slides of paper presentations, commentaries and of group discussions are being uploaded on the page containing the materials of the seminar.

Given that the Dagstuhl seminar format allowed more time and flexibility than a regular workshop, we were able to invite a participation of the European project COST Action "Agreement Technologies" (www.agreement-technologies.eu), that has Norms in Multi Agent Systems as the topic of one of its five working groups, of which many of the NorMas-09 participants are members or supporters. The COST action organized a panel discussion on normative aspects of MAS in which the working group leader Cristiano Castelfranchi was accompanied by Thomas Agotnes, Guido Boella, Pablo Noriega and Timothy Norman. The COST action contributed further to the seminar by funding the participation of Bastin Savarimutu, a young researcher from University of Otago in New Zealand.

In addition to the presentations, group discussions and the Agreement Technologies panel, the seminar hosted an evening session about projects on Normative Multi Agent Systems where participants are currently involved. Although a similar initiative had taken place at NorMAS Dagstuhl Seminar 2007, the situation was then still too preliminary. In this session, the survey of projects showed that the area is in a healthy development. More than 15 projects are being funded at different levels: from international (ITA), European (ALIVE, COST Action Agreement technologies) as well as national and regional ones (ICT4LAW, Agreement technologies, Llibre Blanc de la Mediacio en Catalunya). A page of the WIKI website will include links to such projects. The session served also to discuss project proposals. From the discussion it became apparent that the ICT call of FP7 does not seem to hold a clearly defined spot for NorMAS topics. However, a new opportunity was identified: presenting an EUROCORE theme proposal. The goal of this initiative is to show to the Commission the existence of a large community working on norms: 44 researches participated to the seminar, but more than 80 people had been invited, and many of them have large groups of people working also on norms.

Since the community has very different lines of research and to an extent belongs to different disciplines like sociology, law, computer science, economics, one of the goal of the seminar was to strengthen the roots of the community. For this reason, and in line with the COST Action Agreement technologies supporting the seminar, we agreed to identify the most significant papers for the NorMAS community with the likely intent to produce a seminal book around them. Thus, the participants have been invited to send a list of the papers which they believe to be most significant for normative multiagent systems and the organizers will sort out the results to get this initiative going.

As with the previous Dagstuhl seminar on NorMAS, the best revised papers will be published in specialized journals. In this occasion, a special issue of the Journal of Algorithms in Cognition, Informatics and Logic of Elsevier (ISI impact factor 1.2, A class in the Italian GRIN rating) and a special issue of the Artificial Intelligence and Law Journal will be produced after a new round of anonymous peer reviews.

The interest of the community present at Dagstuhl seems to have shifted from applications like Second Life to other domains like normative compliance, norm detection and norm enforcement. The amount of application papers was a bit higher than at NorMAS07 at Dagstuhl, partially due to new research groups joining the community to present their work and the interest of Web2.0 based applications in the area of norms.

New methodologies have been introduced like argumentation theory, while other more traditional ones like deontic logic have reduced their importance. Still some uncertainty exists about which is the right methodology: for example traditional modal approaches have show some limitations. The meta-question being "is there a "right" methodology?" probably has the answer "no, it depends on the research question and focus". We are happy to see an increasing spread in the type of questions and answers being put forward within the community, while at the same time the discussions suggest that the shared vocabulary and views do not suffer from this.

We are planning to continue on the NorMAS journey by organizing NorMAS2010 (probably at AISB2010) and NorMAS2011 (at Dagstuhl preferably) and setting up a NorMAS steering group.