Sir — Your News Feature “Save our city!”(Nature 424, 608–609; 2003) reviews thedilemma faced in selecting a large-scaleengineering project to protect Venice, to be implemented and financed by theItalian government. I can see benefits in both of the two major projectsdiscussed: the MOSE series of moveablegates, and the scheme to separate the cityfrom its lagoon and the sea.Unfortunately, the first option is likely to serve only as a temporary protectivemeasure for a century or two and causecollateral problems, such as increasedpollution. The second, on the other hand, will permanently alter theconfiguration and setting of the Venice

How do impact factorsrelate to the real world?Sir — I have read with interest your debateon the impact of scientific work (forexample, Nature 422, 259–261; 2003,Nature 423, 479–480 & 585; 2003 andNature 424, 14; 2003) but I do not agreewith the position taken by Adam /Lomnicki(Nature 424, 487; 2003).

The problem is that /Lomnicki andothers established their careers at a timewhen competition among scientists had a different meaning. I am a young scientist and like everyone I would like to discover something interesting and new. However, when my colleagues and I discuss biological problems we alwaysthink about impact factors.

Recently, my friends wondered where to send their new paper — to journal Xwith an impact factor of 1.4, or to journal Y with an impact factor of 1.8. After 10years, the average paper will be cited 14times in journal X and 18 times in journalY (assuming a constant citation rate, which is not the case, of course). Thus wecompete furiously for just a few morecitations, as the impact factor of mostjournals does not exceed three.

As /Lomnicki states, citations arestatistical processes, but even very goodpapers are cited only a few times. Thequestion is whether the difference between 10 and 20 citations can reallychange our knowledge and understandingof nature.

Journals as well as scientists competefor impact factors. A journal that wants a higher impact factor has to encourageauthors to publish in it, but with morepapers coming in, more must be rejected.Generally, authors want to publish in

we know, a city that is charmed andrecognized by its lagoon.

Various ancient Greek coastal cities,some with canals and constructedcenturies earlier than Venice, are nowcompletely submerged, such as Herakleion off the coast of Egypt. They,too, faced problems of insufficientprotection against water surges and sea-level rise, and subsidence caused by the building of monumental structures on inadequate foundations. Towards the end of their active history, when the sea level rose above the base of the buildings, structures toppled increasinglyrapidly until the cities were finallysubmerged.

journals with the highest possible impactfactor, but it is very difficult for mostjournals to improve their impact factors as most submitted papers have beenrejected by better journals.

Impact factors provide an easy way toassess our achievements. But we do notknow if small differences in citationnumber are valid indicators of our work,or how citations are related to the realworld and solving its problems.Piotr Skórka Institute of Environmental Sciences,Jagiellonian University, ul. Gronostajowa 3,30-387 Kraków, Poland

Pointless suffering ofanimals can be avoidedSir — In the News article “Agony forresearchers as mix-up forces retraction of ecstasy study” (Nature 425, 109; 2003)the ‘agony’ of the embarrassed researcherswas dwarfed by that of their primatesubjects.

From an animal-welfare standpoint,the affair was tragic. From a scientificstandpoint, it was pointless. After all, theresearchers themselves have noted thatextensive evidence from animal studiesalready shows that methylenedioxy-methamphetamine (MDMA, or ecstasy) is dangerous. If more proof of itsrecreational effects is needed, scientistsshould focus on the many humans who use such drugs.

Ethical and non-invasive studies areeasily conducted and, in fact, are alreadybeing carried out to allow scientists todetect early signs of Parkinson’s disease or other neurological effects in drug users.

correspondence

NATURE | VOL 425 | 16 OCTOBER 2003 | www.nature.com/nature 661

In geological terms, these cities weresimilar to Venice. Considering what befell them, the opposite of the geologicaltenet “the present is key to the past” maywell apply to Venice. Perhaps a longer-term, more secure approach would be one involving comprehensive Netherlands-type ‘polder’ dykeconstructions, with seawater beingpumped out of the encircled city, watermaintained in the canals, and those unique structures reinforced, wherepossible, by deep pilings.Jean-Daniel StanleyGeoarchaeology Program, National Museum ofNatural History, Smithsonian Institution,Washington DC 20560, USA

Effective protection may alter the look of VeniceThose seeking to rescue Venice are caught between the devil and the deep-blue sea.

In research currently being conducted by J. H. Atkinson at the University ofCalifornia, San Diego, human users ofmethamphetamine (speed) — the drugaccidentally used in the ‘ecstasy study’ —are examined for neurological andcognitive function and other clinicalfeatures. A similar strategy would apply to the study of MDMA.Neal D. BarnardPhysicians Committee for Responsible Medicine,5100 Wisconsin Avenue, Suite 400,Washington DC 20016, USA

JET at risk if Europe cannot afford to pay for ITERSir — I agree with much of what RichardButtery says in his letter (Nature 424, 995;2003) about the Joint European Torus(JET), but I must point out that he hasquoted my statement (Nature 424, 4;2003) out of context. I was asked by Nature to comment on the situation thatmight occur if there were no substantialincrease in the fusion budget in theSeventh Euratom Framework Programme.I mentioned the possible closure of JET in the context of there being insufficientmoney to finance the Europeancontribution to ITER.A. M. BradshawMax-Planck-Institut für Plasmaphysik,Boltmannstrasse 2, Garching/Greifswald, Germany

correspondenceContributions to Correspondence may be submitted tocorres@nature.com. They should be no longer than500 words, and ideally shorter. Published contributions are edited.

© 2003 Nature Publishing Group