Gott replies: Jorge Hirsch acknowledges his h-index cannot be used to compare people across epochs. That’s where I started. Werner Marx and Manuel Cardona think people can be compared across epochs, and they and Lutz Bornmann propose a “renormalized” h-index to do it.1 

But even comparing contemporaries, Hirsch notes,

For an author with a relatively low h that has a few seminal papers with extraordinarily high citation counts, the h index will not fully reflect that scientist’s accomplishments. Conversely, a scientist with a high h achieved mostly through papers with many coauthors would be treated overly kindly by his or her h.2 

My E-index addresses those problems by weighting all papers in proportion to their citation impact and dividing credit for each paper among its n authors: counting 1⁄2 first-author citations and 1⁄2 fractional (1⁄n) citations. Vladimir Krasnopolsky (PHYSICS TODAY, September 2004, page 12) proposed giving 1⁄2 citation to the first author of a multiauthor paper with the remaining 1⁄2 divided between coauthors—close to what I am doing.

Marx and Cardona actually favor including informal last-name citations, saying,

The data reveal that the formal citations often measure only a small fraction of the overall impact of seminal publications. Furthermore, informal citations are mainly given instead of (and not in addition to) formal citations. As a major consequence, the overall impact of pioneering articles and researchers cannot be entirely determined by merely counting the full reference based citations.3 

The E-index includes informal citations by adding last-name citations in titles and abstracts to the average of first-author and fractional citations. That recovers additional “lost” citations that Albert Einstein and other greats are getting, mostly from recent times, and allows comparison with recent physicists without requiring renormalization.

The e-index that Cardona and Marx mention is for excess citations4 and is not to be confused with my E-index. It recognizes the seminal-paper problem by counting all citations in the author’s most cited h papers to come closer to the total citation count.

The E-index is proportional to impact and can be measured in milli-Einsteins (mE). C. V. Raman is high, but when people mention Raman spectroscopy in titles or abstracts, it’s hard to argue they aren’t referring to his Nobel Prize–winning work. The 20th- and 21st-century physicists and astronomers I’ve found with the top E-index values are Raman (1468 mE), Enrico Fermi (1277), Einstein (1000), and Edwin Hubble (815), using the SAO/NASA Astrophysics Data System (ADS) database. (On ADS, Einstein’s h-index was 27, as I reported, and is now 28. On the larger ISI database covering all science, it is 50.) Time magazine selected Einstein, Fermi, Hubble, and William Shockley as the most influential 20th-century physicists and astronomers, three out of four in common with the E-index.

The E-index’s linear scale allows comparison of magnitudes of impact, something that Lev Landau was interested in.5 He estimated that physicists of class 1, like Werner Heisenberg, made 10–0.5 of Einstein’s contribution (or an E-index of 316 mE, close to Heisenberg’s E-index of 417 mE); class 2, 10–1.5; class 3, 10–2.5; and class 4, 10–3.5. Landau had a class 5, for those making negative contributions—where Ravi Rau clearly thinks those working on citation indices belong!

1.
W.
Marx
,
L.
Bornmann
,
M.
Cardona
,
J. Am. Soc. Inf. Sci. Tech.
61
,
2061
(
2010
).
2.
J. E.
Hirsch
,
Proc. Natl. Acad. Sci. USA
102
,
16569
(
2005
).
3.
W.
Marx
, M. Cardona,
Scientometrics
80
,
1
(
2009
).
4.
C. -T.
Zhang
,
PLoS ONE
4
(
5
),
e5429
(
2009
).
5.
E. M.
Lifshitz
, in
Mechanics
, 3rd ed.,
L. D.
Landau
,
E. M.
Lifshitz
,
Butterworth-Heinemann
,
Oxford, UK
(
1976
), p.
xvii
.