Modular invariance in non-rational CFT

Modular invariance of the torus partition function is often the first -- and sometimes the only -- thing people check about a proposed CFT. There is a good reason for this: computing the torus partition function only requires knowing the characters of the representations which appear in the spectrum, whereas other consistency checks, such as crossing symmetry of the sphere four-point function, involve much more complicated conformal blocks.

However, modular invariance is neither sufficient, nor necessary for a CFT to be consistent.

Les critères d'inscription à l'HDR en Physique à l'université Paris-Sud

Bien que le dossier d'inscription à l'HDR soit décrit en détail sur la page de l'université, rien n'indique selon quels critères un candidat sera autorisé à s'inscrire ou non. Le dossier comprend naturellement une liste de publications et des indications sur les étudiants déjà suivis, mais combien faut-il de publications? et quelle expérience d'encadrement est nécéssaire? J'ai reçu récemment quelques explications à ce sujet de Réza Ansari, le correspondant HDR pour la physique théorique entre autres.

Bourses doctorales de l'ED PIF: quelques précisions

L'École Doctorale Physique en Île-de-France (ED PIF) vient de procéder à son premier concours de recrutement des doctorants. Ayant présenté un candidat (malheureux) à ce concours, et ayant eu des échanges avec certains membres du jury, je peux donner quelques éléments sur le fonctionnement du concours, dans le but d'informer les candidats potentiels aux prochaines éditions.

Positive feedback loops in research funding

Funding research should mean sending money where it is most useful, considering the strength of the researchers to be funded, the interest of their projects, and how much these projects need money. However, in practice, already being well-funded or belonging to a well-funded institution is helpful, and sometimes even necessary, for a researcher to obtain funding. So there are positive feedback loops in research funding, which should be distinguished from other mechanisms that tend to concentrate research funding -- such as some people being better than others at getting money. (That these people are not necessarily the best researchers is not our subject here.)

Let's discuss some positive feedback loops:

Le boeuf qui voulait se faire aussi petit qu'une grenouille

Je propose ici quelques commentaires sur la stratégie du CNRS pour l'information scientifique et technique, telle qu'exposée dans un document de 133 pages. Le titre de mon texte exprime le constat qu'avec cette stratégie, le CNRS ne semble pas vouloir se comporter en gros acteur institutionnel capable d'influencer le comportement des nombreux chercheurs et laboratoires qui dépendent de lui, mais plutôt en prestataire de services qui se contente de les informer et de leur proposer des outils.

La stratégie en matière de libre accès (Open Access) illustre bien ce constat.

Rating scientific articles: why and how?

Assuming we need to assess the quality of individual scientific articles, there are two broad approaches:

• the quantitative approach of counting citations and/or computing alternative metrics,
• the qualitative approach of conducting peer review, and giving quality tokens to articles. A quality token can be publication in a journal, or a grade.

The quantitative approach has the fundamental flaw of being an indirect way of assessing quality. And it has undesirable side effects, such as the misuse of citations for assigning credit, rather than for helping readers. In this post I will deal with the qualitative approach.

Giving grades, implicit or explicit.

Assuming some form of peer review is conducted, how can the results be concisely summarized?

Bureaucratie contre recherche: quelques nouvelles du front [Bureaucracy vs research: news from the frontline]

[Ajout le 3 avril 2014:] L'ED PIF a maintenant un site web!

[Ajout le 25 mars 2014:] L'ED PIF distribue maintenant par email son guide du doctorant.

Comme de nombreux collègues, j'ai récemment eu connaissance de la nouvelle procédure de recrutement des doctorants de l'ED PIF (École doctorale Physique en Île-de-France), décrite dans ce document PDF. Cette ED PIF vient de naître de la fusion de plusieurs Écoles doctorales. Tout étudiant souhaitant faire une thèse en physique devra en obtenir l'approbation préalable. Le but affiché de la procédure est d'aider étudiants et chercheurs à faire les bons choix quand ils élaborent ensemble des projets de thèses, afin de réduire autant que possible les erreurs et les échecs.

Cherche thèse d'urgence.

Or, par rapport à ce but, la procédure proposée pourrait bien être contre-productive. En effet, les étudiants qui demanderont une bourse de l'École doctorale devront choisir leur sujet de thèse avant le 28 avril, et défendre leur candidature dans un concours avec dossier et audition.

The case for emancipating articles from their authors

In August 1795, the composer Joseph Haydn, on his way back from England, stopped at Passau, where he listened to a performance of one of his most famous works, the Seven Last Words of Christ. He had written this work in 1786, at the request of a Spanish canon, who wanted orchestral music for accompanying meditation during church services. At Passau however, the local Kapellmeister, Joseph Friebert, had added a choir, and made the work into a cantata. Haydn himself had already done a transcription for string quartet, and approved another transcription for piano -- but he had never thought about a cantata.

So, how did Haydn react, after his work was modified without his authorization?

The world according to Elsevier

Elsevier is considered by some academics as the worst predatory publisher, and is a favourite target for boycotts. In particular, the negotiations for subscriptions to bundles of journals, which were the subject of my previous post, are particularly difficult in the case of Elsevier, leading to much frustration for academic institutions. But what sets Elsevier apart from other academic publishers? Let me give some tentative answers, based on observations on Elsevier's behaviour. I will first give the data, and then try to explain them in terms of a coherent strategy.

1. Elsevier considers open access to scientific publications as a threat, and fights against it. This is explicitly said (in milder terms) in their 2011 financial review.

Imminent capitulation of research institutes in "negotiations" with publishers

Recently, I have read pathetic emails from librarians who are involved in "negotiations" with publishers for the renewal of subscriptions to academic journals -- subscriptions which, if not renewed, will lapse at the end of this year. Subscriptions are "negotiated" at a high level: nationwide consortiums of research institutes on one side, large publishers selling bundles of hundreds of journals on the other side. Here is how such "negotiations" go: the publisher demands an outrageous price, and the consortium agrees to that price, minus a symbolic discount if the publisher wants the consortium to save face.

This outcome is the inevitable result of the basic economics of the game. A big publisher enjoys a de facto monopoly on his journals, and can therefore set prices based on what they think the consortium can pay. The only way out for the consortium would be to threaten not to subscribe. But the larger the consortium, the more researchers there are who will say they cannot live without the subscription, and the least credible the threat is. Anyway, consortiums are built with the specific aim of subscribing, and do not have the option to do otherwise.

This is why the "negotiations" with publishers are not really negotiations, and why the negotiators are so dispirited.

Bibcac: a Perl script for classifying references in a Bibtex bibliography

In my previous post I proposed a script for updating a Bibtex file with publication data. Here is now a completely independent script for organizing the references in a Bibtex file.

To do this, we will not modify the Bibtex file, but rather create an auxiliary file where we will associate categories and comments to a number of the references. For instance, if we want to associate the category "Topological Recursion" and the comment "This article..." to the article whose Bibtex key is "cer12", we will add the following text in the auxiliary file:

\ArticleLabel{cer12}
\ArticleCategory{Topological Recursion}

This article ...

Why would we want to have such categories and comments? Categories are particularly useful if they match some classification of the articles, whether as files in a computer, or as physical printouts. If you keep articles in a number of folders, the program will enable you to find in which folder a given article is, and to list the articles in a given folder. As for comments, keeping them in a centralized file is safer than having them as annotations (handwritten or electronic) on a copy of an article, which might well get lost.

Bibpup: a Perl script for updating Bibtex files using Inspire

Bibtex is a nice way to manage bibliographies, by collecting the bibliographical data of many articles in one Bibtex file, and then citing only some of these articles in a given document. However, the bibliographical data of an article can change when the article gets published in a journal. If the article was initially entered in the Bibtex file as a preprint, the Bibtex file must then be updated.
Here I propose a Perl script which does this automatically using the search engine Inspire, which is the standard search engine for high-energy physics and related fields.

Of course I do not believe that it is very relevant or useful to know whether an article is published and in which journal, but journals typically require such data to be displayed in bibliographies. In any case, the decision to display publication data or not is done at the level of the bibliography style file -- the Bibtex file itself is only a database, which should be as complete as possible.

So what does the script do? the script takes a Bibtex file, and for each entry which has no publication data, but an eprint field with an Arxiv number, a request is sent to Inspire. If publication data are found, then the entry is modified to include them. (The rest of the entry, and in particular the key, are not changed.)