Papers and patents are becoming less disruptive over time

By Michael Park, Erin Leahey & Russell J. Funk

Annotated on Fermat's Library with 8 comments. Text and annotations as Markdown Original paper

This paper made quite a splash when it came out. This Nature editor...
In 1675, Isaac Newton remarked in a letter to rival Robert Hooke: “...
>> “We find that papers and patents are increasingly less likely to...
>>"Yet much remains unknown, not merely about the causes of slowing...
The analyses and methods shared in this paper hinge on the CD metri...
Even if the rates of disruptive work are going down, the absolute c...
Interesting to see that Watson and Crick's DNA discovery isn't even...
>> "Overall, our results deepen understanding of the evolution of k...
138 | Nature | Vol 613 | 5 January 2023
Article
Papers and patents are becoming less
disruptive over time
Michael Park
1
, Erin Leahey
2
& Russell J. Funk
1
 ✉
Theories of scientic and technological change view discovery and invention as
endogenous processes
1,2
, wherein previous accumulated knowledge enables future
progress by allowing researchers to, in Newton’s words, ‘stand on the shoulders of
giants’
3–7
. Recent decades have witnessed exponential growth in the volume of new
scientic and technological knowledge, thereby creating conditions that should be
ripe for major advances
8,9
. Yet contrary to this view, studies suggest that progress is
slowing in several major elds
10,11
. Here, we analyse these claims at scale across six
decades, using data on 45 million papers and 3.9 million patents from six large-scale
datasets, together with a new quantitative metric—the CD index
12
—that characterizes
how papers and patents change networks of citations in science and technology.
We nd that papers and patents are increasingly less likely to break with the past in
ways that push science and technology in new directions. This pattern holds universally
across elds and is robust across multiple dierent citation- and text-based metrics
1,13–17
.
Subsequently, we link this decline in disruptiveness to a narrowing in the use of previous
knowledge, allowing us to reconcile the patterns we observe with the ‘shoulders of
giants’ view. We nd that the observed declines are unlikely to be driven by changes in
the quality of published science, citation practices or eld-specic factors. Overall,
our results suggest that slowing rates of disruption may reect a fundamental shift in
the nature of science and technology.
Although the past century witnessed an unprecedented expan-
sion of scientific and technological knowledge, there are concerns
that innovative activity is slowing
18–20
. Studies document declining
research productivity in semiconductors, pharmaceuticals and other
fields
10,11
. Papers, patents and even grant applications have become less
novelrelative to prior work and less likely to connect disparate areas
of knowledge, both of which are precursors of innovation
21,22
. The gap
between the year of discovery and the awarding of a Nobel Prize has also
increased
23,24
, suggesting that today’s contributions do not measure
up to the past. These trends have attracted increasing attention from
policymakers, as they pose substantial threats to economic growth,
human health and wellbeing, and national security, along with global
efforts to combat grand challenges such as climate change
25,26
.
Numerous explanations for this slowdown have been proposed.
Some point to a dearth of ‘low-hanging fruit’ as the readily available
productivity-enhancing innovations have already been made
19,27
. Others
emphasize the increasing burden of knowledge; scientists and inven-
tors require ever more training to reach the frontiers of their fields,
leaving less time to push those frontiers forward
18,28
. Yet much remains
unknown, not merely about the causes of slowing innovative activity,
but also the depth and breadth of the phenomenon. The decline is
difficult to reconcile with centuries of observation by philosophers of
science, who characterize the growth of knowledge as an endogenous
process, wherein previous knowledge enables future discovery, a view
captured famously in Newton’s observation that if he had seen further,
it was by ‘standing on the shoulders of giants’
3
. Moreover, to date, the
evidence pointing to a slowdown is based on studies of particular fields,
using disparate and domain-specific metrics
10,11
, making it difficult to
know whether the changes are happening at similar rates across areas
of science and technology. Little is also known about whether the pat-
terns seen in aggregate indicators mask differences in the degree to
which individual works push the frontier.
We address these gaps in understanding by analysing 25 million
papers (1945–2010) in the Web of Science (WoS) (Methods) and 3.9 mil-
lion patents (1976–2010) in the United States Patent and Trademark
Office’s (USPTO)Patents View database (Methods). The WoS data
include 390 million citations, 25 million paper titles and 13 million
abstracts. The Patents View data include 35 million citations, 3.9 million
patent titles and 3.9 million abstracts. Subsequently, we replicate our
core findings on four additional datasets—JSTOR, the American Physical
Society corpus, Microsoft Academic Graph and PubMed—encompass-
ing 20 million papers. Using these data, we join a new citation-based
measure
12
with textual analyses of titles and abstracts to understand
whether papers and patents forge new directions over time and across
fields.
Measurement of disruptiveness
To characterize the nature of innovation, we draw on foundational
theories of scientific and technological change
2,29,30
, which distinguish
between two types of breakthroughs. First, some contributions improve
existing streams of knowledge, and therefore consolidate the status
https://doi.org/10.1038/s41586-022-05543-x
Received: 14 February 2022
Accepted: 8 November 2022
Published online: 4 January 2023
Check for updates
1
Carlson School of Management, University of Minnesota, Minneapolis, MN, USA.
2
School of Sociology, University of Arizona, Tucson, AZ, USA.
✉
e-mail: rfunk@umn.edu
Nature | Vol 613 | 5 January 2023 | 139
quo. Kohn and Sham (1965)
31
, a Nobel-winning paper used established
theorems to develop a method for calculating the structure of elec-
trons, which cemented the value of previous research. Second, some
contributions disrupt existing knowledge, rendering it obsolete, and
propelling science and technology in new directions. Watson and Crick
(1953)
32
, also a Nobel winner, introduced a model of the structure of
DNA that superseded previous approaches (for example, Pauling’s
triple helix). Kohn and Sham and Watson and Crick were both impor-
tant, but their implications for scientific and technological change
were different.
We quantify this distinction using a measure—the CD index
12
—that
characterizes the consolidating or disruptive nature of science and
technology (Fig.1). The intuition is that if a paper or patent is disrup-
tive, the subsequent work that cites it is less likely to also cite its prede-
cessors; for future researchers, the ideas that went into its production
are less relevant (for example, Pauling’s triple helix). If a paper or patent
is consolidating, subsequent work that cites it is also more likely to
cite its predecessors; for future researchers, the knowledge upon
which the work builds is still (and perhaps more) relevant (for exam-
ple, the theorems Kohn and Sham used). The CD index ranges from −1
(consolidating) to 1 (disruptive). We measure the CD index five years
after the year of each paper’s publication (indicated by CD
5
, see
Extended Data Fig.1 for the distribution of CD
5
among papers and
patents and Extended Data Fig.2 for analyses using alternative
windows)
33
. For example, Watson and Crick and Kohn and Sham both
received over a hundred citations within five years of being published.
However, the Kohn and Sham paper has a CD
5
of −0.22 (indicating
consolidation), whereas the Watson and Crick paper has a CD
5
of
0.62 (indicating disruption). The CD index has been validated exten-
sively in previous research, including through correlation with expert
assessments
12,34
.
Declining disruptiveness
Across fields, we find that science and technology are becoming less
disruptive. Figure2 plots the average CD
5
over time for papers (Fig.2a)
and patents (Fig.2b). For papers, the decrease between 1945 and 2010
ranges from 91.9% (where the average CD
5
dropped from 0.52 in 1945
to 0.04 in 2010 for ‘social sciences’) to 100% (where the average CD
5
decreased from 0.36 in 1945 to 0 in 2010 for ‘physical sciences’); for pat-
ents, the decrease between 1980 and 2010 ranges from 78.7% (where the
average CD
5
decreased from 0.30 in 1980 to 0.06 in 2010 for ‘computers
and communications’) to 91.5% (where the average CD
5
decreased from
0.38 in 1980 to 0.03 in 2010 for‘drugs and medical’). For both papers
and patents, the rates of decline are greatest in the earlier parts of the
time series, and for patents, they appear to begin stabilizing between
the years 2000 and 2005. For papers, since about 1980, the rate of
decline has been more modest in ‘life sciences and biomedicine’ and
physical sciences, and most marked and persistent in social sciences
and ‘technology’. Overall, however, relative to earlier eras, recent papers
and patents do less to push science and technology in new directions.
The general similarity in trends we observe across fields is noteworthy
in light of ‘low-hanging fruit’ theories
19,27
, which would probably predict
greater heterogeneity in the decline, as it seems unlikely fields would
‘consume’ their low-hanging fruit at similar ratesortimes.
Linguistic change
The decline in disruptive science and technology is also observable
using alternative indicators. Because they create departures from
the status quo, disruptive papers and patents are likely to introduce
new words (for example, words used to create a new paradigm might
differ from those that are used to develop an existing paradigm)
35,36
.
bc
Disruptive
Consolidatin
g
CD
t
= 1
a
–2(1)(0) + (1) = 1
CD
t
=
(0) + (-1) + (1) + (1) + (–1) + (0) = 0
Focal paper/patent’s precedecessor
CD
t
= 0
CD
t
= –1
Papers Patents
WoS USPTO
Maximally disruptive
CD
t
= 1
Time
–2(1)(1) + (1) = –1
–2(0)(1) + (0) = 0
–2(0)(1) + (0) = 0
–2(1)(1) + (1) = –1
–2(1)(0) + (1) = 1
CD
t
=
1
–
n
Σ
–2f
it
b
it
+ f
it
n=i
n
CD
5
= 0.70
(Axel, Wigler and Silverstein 1983): Patent on
eukaryotic cotransformation, which proposed
inserting foreign genes into cells that then
produce associated proteins; foundational
method in biotechnology.
CD
5
= 0.62
(Watson and Crick 1953): Nobel-prize-
winning paper that proposed a new structure
of DNA; debunked the existing model of
DNA structure, Pauling’s 3D double helix.
CD
5
= 0.16
(Page 1997): Patent for Google’s core
algorithm to rank the importance of web
pages. Presented a novel method
that utilizes network theory to rank web
pages based on links from other sites.
CD
5
= –0.22
(Kohn and Sham 1965): Nobel-Prize-winning
paper that developed the Kohn–Sham
equation. The equation was an application
of the existing Hohenberg–Kohn theorem.
CD
5
= –0.55
(Baltimore 1970): Nobel-Prize-winning paper
that showed that viruses with RNA
genomes can be inserted into host cells;
developed from existing ndings on the
transmission of genetic information between
DNA and RNA.
CD
5
= –0.85
(Hawbaker 2005): Monsanto’s patent on
a geneticallymodied soybean that is
resistant to glyphosate; integrates other
desirable plant characteristics such as
higher yield, immunity to many diseases,
and resistance to shattering.
Maximally consolidating
CD
t
= –1
CD
t
= 0
Time
Time
f
it
= 1 if i cites the focal paper/patent; 0 if not
b
it
= 1 if i cites the predecessors of the focal
paper/patent; 0 if not
n
t
: number of forward cites to the focal work
and/or its precedessors as of time t
Equation
Cites focal paper/patent
and predecessor (type f = 1, b = 1)
Cites predecessor (type b = 1)
Cites focal paper/patent (type f = 1)
Focal paper/patent
Fig. 1 | Overview of the measurement approach. This figure shows a schematic
visualization of the CD index. a, CD index value of three Nobel Prize-winning
papers
31,32,58
and three notable patents
59–61
in our sample, measured as of five
years post-publication (indicated by CD5). b, Distribution of CD
5
for papers
from WoS (n = 24,659,076) between 1945 and 2010 and patents from Patents
View (n = 3,912,353) between 1976 and 2010, where a single dot represents a
paper or patent. The vertical (up–down) dimension of each ‘strip’ corresponds
to values of the CD index (with axis values shown in orange on the left).
The horizontal (left–right) dimension of each strip helps tominimize
overlapping points. Darker areas on each strip plot indicate denser regions
of the distribution (that is, more commonly observed CD
5
values). Additional
details on the distribution of the CD index are given in Extended Data Fig.1.
c, Three hypothetical citation networks, where the CD index is at the maximally
disruptive value (CD
t
 = 1), midpoint value (CD
t
 = 0), and maximally consolidating
value (CD
t
 = −1). The panel also provides the equation for the CD index and an
illustrative calculation.
140 | Nature | Vol 613 | 5 January 2023
Article
Therefore, if disruptiveness is declining, we would expect a decline
in the diversity of words used in science and technology. To evaluate
this, Fig.3a,d documents the type-token ratio (that is, unique/total
words) of paper and patent titles over time (SupplementaryInforma-
tion section1). We observe substantial declines, especially in the ear-
lier periods, before 1970 for papers and 1990 for patents. For paper
titles (Fig.3a), the decrease (1945–2010) ranges from 76.5% (social
sciences) to 88% (technology); for patent titles (Fig.3d), the decrease
(1980–2010) ranges from 32.5% (chemical) to 81% (computers and
communications). For paper abstracts (Extended Data Fig.3a), the
decrease (1992–2010) ranges from 23.1% (life sciences and biomedicine)
to 38.9% (social sciences); for patent abstracts (Extended Data Fig.3b),
the decrease (1980–2010) ranges from 21.5% (mechanical) to 73.2%
(computers and communications). In Fig.3b,e, we demonstrate that
these declines in word diversity are accompanied by similar declines in
combinatorial novelty; over time, the particular words that scientists
and inventors use in the titles of their papers and patents are increas-
ingly likely to have been used together in the titles of previous work.
Consistent with these trends in language, we also observe declining nov-
elty in the combinations of previous work cited by papers and patents,
based on a previously established measure of ‘atypical combinations’
14
(Extended Data Fig.4).
The decline in disruptive activity is also apparent in the specific
words used by scientists and inventors. If disruptiveness is declining,
we reasoned that verbs alluding to the creation, discovery or percep-
tion of new things should be used less frequently over time, whereas
verbs alluding to the improvement, application or assessment of exist-
ing things may be used more often
35,36
. Figure3 shows the most com-
mon verbs in paper (Fig.3c) and patent titles (Fig.3f) in the first and
last decade of each sample (Supplementary Information section2).
Although precisely and quantitatively characterizing words as ‘con-
solidating’ or ‘disruptive’ is challenging in the absence of context, the
figure highlights a clear and qualitative shift in language. In the earlier
decades, verbs evoking creation (for example, ‘produce’, ‘form’, ‘pre-
pare’ and ‘make’), discovery (for example, ‘determine’ and ‘report’)
and perception (for example, ‘measure’) are prevalent in both paper
and patent titles. In the later decades, however, these verbs are almost
completely displaced by those tending to be more evocative of the
improvement (for example, ‘improve’, ‘enhance’ and ‘increase’), appli-
cation (for example, ‘use’ and ‘include’) or assessment (for example,
‘associate’, ‘mediate’ and ‘relate’) of existing scientific and technologi-
cal knowledge and artefacts. Taken together, these patterns suggest a
substantive shift in science and technology over time, with discovery
and invention becoming less disruptive in nature, consistent with our
results using the CD index.
Conservation of highly disruptive work
The aggregate trends we document mask considerable heterogeneity
in the disruptiveness of individual papers and patents and remarkable
stability in the absolute number of highly disruptive works (Methods
and Fig.4). Specifically, despite large increases in scientific productiv
-
ity, the number of papers and patents with CD
5
values in the far right tail
of the distribution remains nearly constant over time. This ‘conserva-
tion’ of the absolute number of highly disruptive papers and patents
holds despite considerable churn in the underlying fields responsible
for producing those works (Extended Data Fig.5, inset). These results
suggest that the persistence of major breakthroughs—for example,
measurement of gravity waves and COVID-19 vaccines—is not incon-
sistent with slowing innovative activity. In short, declining aggregate
disruptiveness does not preclude individual highly disruptive works.
Alternative explanations
What is driving the decline in disruptiveness? Earlier, we suggested our
results are not consistent with explanations that link slowing innova-
tive activity to diminishing ‘low-hanging fruit’. Extended Data Fig.5
shows that the decline in disruptiveness is unlikely to be due to other
field-specific factors by decomposing variation in CD
5
attributable to
field, author and year effects (Methods).
Declining rates of disruptive activity are unlikely to be caused by the
diminishing quality of science and technology
22,37
. If they were, then
the patterns seen in Fig.2 should be less visible in high-quality work.
However, when we restrict our sample to articles published in premier
publication venues such as Nature, Proceedings of the National Academy
of Sciences and Science or to Nobel-winning discoveries
38
(Fig.5), the
downward trend persists.
Furthermore, the trend is not driven by characteristics of the WoS and
UPSTO data or our particular derivation of the CD index; we observe
similar declines in disruptiveness when we compute CD
5
on papers
1950 1960 1970 1980 1990 2000 2010
Year
0
0.1
0.2
0.3
0.4
0.5
a
Papers
Life sciences and biomedicine
Physical sciences
Social sciences
Technology
1980 1990 2000
2010
Year
0
0.1
0.2
0.3
0.4
0.5
Average CD
5
Patents
Chemical
Computers and communications
Drugs and medical
Electrical and electronic
Mechanical
b
Average CD
5
Fig. 2 | Decline of disruptive science and technology. a,b, Decline in CD
5
over
time, separately for papers (a, n = 24,659,076) and patents (b, n = 3,912,353).
For papers, lines correspond to WoS research areas; from 1945 to 2010 the
magnitude of decline ranges from 91.9% (social sciences) to 100% (physical
sciences). For patents, lines correspond to National Bureau of Economic
Research (NBER) technology categories; from 1980 to 2010 the magnitude
of decline ranges from 93.5% (computers and communications) to 96.4%
(drugs and medical). Shaded bands correspond to 95% confidence intervals.
As we elaborate in the Methods, this pattern of decline is robust to adjustment
for confounding from changes in publication, citation and authorship practices
over time.
Nature | Vol 613 | 5 January 2023 | 141
1950 1960 1970 1980 1990 2000 2010
Year
0
0.05
0.10
0.15
0.20
0.25
0.30
0.35
Unique words/total words
a
Life sciences and biomedicine
Physical sciences
Social sciences
Technology
1950 1960 1970 1980 1990 2000 2010
Year
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
New word pairs/total word pairs
b
1980 1990 2000 2010
Year
0
0.05
0.10
0.15
0.20
0.25
0.30
0.35
Unique words/total words
d
Chemical
Computers and communications
Drugs and medical
Electrical and electronic
Mechanical
1980 1990 2000 2010
Year
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
New word pairs/total word pairs
e
Verb use per 100 verbs
follow 0.60
treat 0.44
produce 0.32
report 0.32
determine 0.25
affect 0.24
measure 0.23
note 0.16
inuence 0.16
study
0.14
1950
Verb use per 100 verbs
base 3.42
induce 2.21
associate 1.15
mediate 0.73
improve 0.63
enhance 0.63
control 0.62
relate 0.58
follow 0.56
increase 0.55
2010
Verb use per 100 verbs
have 2.58
contain 2.08
produce 2.00
make 1.94
use 1.47
prepare 1.19
control 1.14
form 1.00
mean 0.94
measure 0.63
1980
Verb use per 100 verbs
use
5.87
have 4.03
form 2.43
base 2.28
manufacture 1.91
control 1.64
produce 1.54
provide 1.36
include 1.35
make 1.14
2010
c
f
Papers
Patents
Fig. 3 | Decline of disruptive science and technology is visible in the
changing language of papers and patents. a,d, Figures showing a decline in
the diversity of language used in science and technology based on the unique/
total words of paper titles from 1945 to 2010 (a, n = 24,659,076) and of patent
titles from 1980 to 2010 (d, n = 3,912,353). b,e, Figures showing a decline in the
novelty of language used in science and technology based on the number of
new word pairs/total word pairs introduced each year in WoS paper titles from
1945 to 2010 (b) and in Patents View patent titles from 1980 to 2010 (refs.
1,17
)(e).
For papers in both a and b, lines correspond to WoS research areas (n = 264 WoS
research area × year observations). For patents in both d and e, lines correspond
to NBER technology categories (n = 229 NBER technology category × year
observations). c,f, Figures showing the frequency of the most commonly used
verbs in paper titles for the first (red) and last (blue) decades of the observation
period in paper (c, n = 24,659,076) and patent (f, n = 3,912,353) titles.
1950 1960 1970 1980 1990 2000 2010
Year
0
25,000
50,000
75,000
100,000
125,000
150,000
175,000
Count of papers
Technology
Social sciences
Physical sciences
Life sciences
a
Papers
Year
Composition of most
disruptive papers by
eld over time
1980 1990 2000 2010
Year
0
10,000
20,000
30,000
40,000
50,000
60,000
Count of patents
MechanicalMechanical
Electrical
Electrical
Drugs
Drugs
Computers
Computers
Chemical
Chemical
b
Patents
(0, 0.25)
(0.25, 0.50)
(0.50, 0.75)
(0.75, 1.00)
CD
5
value
Fig. 4 | Conservation of highly disruptive work. This figure shows the number
of disruptive papers (a, n = 5,030,179) and patents (b, n = 1,476,004) across four
different ranges of CD
5
(papers and patents with CD
5
values in the range [−1.0, 0)
are not represented in the figure). Lines correspond to different levels of
disruptiveness as measured by CD
5
. Despite substantial increases in the numbers
of papers and patents published each year, there is little change in the number of
highly disruptive papers and patents, as evidenced by the relatively flat red, green
and orange lines. This pattern helps to account for simultaneous observations
of both aggregate evidence of slowing innovative activity and seemingly major
breakthroughs in many fields of science and technology. The inset plots show
the composition of the most disruptive papers and patents (defined as those
with CD
5
values >0.25) by field over time. The observed stability in the absolute
number of highly disruptive papers and patents holds despite considerable
churn in the underlying fields of science and technology responsible for
producing those works. ‘Life sciences’ denotes the life sciences and biomedicine
research area; ‘electrical’ denotes the electrical and electronic technology
category; ‘drugs’ denotes the drugs and medical technology category; and
‘computers’ denotes the computers and communications technology category.
142 | Nature | Vol 613 | 5 January 2023
Article
in JSTOR, the American Physical Society corpus, Microsoft Academic
Graph and PubMed (Methods), the results of which are shown in
Extended Data Fig.6. We further show that the decline is not an arte-
fact of the CD index by reporting similar patterns using alternative
derivations
13,15
(Methods and Extended Data Fig.7).
Declines in disruptiveness are also not attributable to changing
publication, citation or authorship practices (Methods). First, using
approaches from the bibliometrics literature
39–43
, we computed several
normalized versions of the CD index that adjusted for the increasing
tendency for papers and patents to cite previous work
44,45
. Results using
these alternative indicators (Extended Data Fig.8a,d) were similar to
those we reported in Fig.2. Second, using regression, we estimated
models of CD
5
as a function of indicator variables for each paper or
patent’s publication year, along with specific controls for field × year
level—number of new papers/patents, mean number of papers/patents
cited, mean number of authors or inventors per paper—and paper or
patent-level—number of papers or patents cited—factors. Predictions
from these models indicated a decline in disruptive papers and patents
(Extended Data Fig.8b,e and Supplementary Table1) that was con-
sistent with our main results. Finally, using Monte Carlo simulations,
we randomly rewired the observed citation networks while preserving
key characteristics of scientists’ and inventors’ citation behaviour,
including the number of citations made and received by individual
papers and patents and the age gap between citing and cited works.
We find that observed CD
5
values are lower than those from the simu-
lated networks (Extended Data Fig.8c,f), and the gap is widening: over
time, papers and patents are increasingly less disruptive than would be
expected by chance. Taken together, these additional analyses indicate
that the decline in CD
5
is unlikely to be driven by changing publication,
citation or authorship practices.
Growth of knowledge and disruptiveness
We also considered how declining disruptiveness relates to the growth
of knowledge (Extended Data Fig.9). On the one hand, scientists and
inventors face an increasing knowledge burden, which may inhibit
discoveries and inventions that disrupt the status quo. On the other
hand, as previously noted, philosophers of science suggest that exist-
ing knowledge fosters discovery and invention
3,6,7
. Using regression
models, we evaluated the relationship between the stock of papers
and patents (a proxy for knowledge) within fields and their CD
5
(Sup-
plementary Information section3 and Supplementary Table2).
We find a positive effect of the growth of knowledge on disruptive-
ness for papers, consistent with previous work
20
; however, we find a
negative effect for patents.
Given these conflicting results, we considered the possibility that
the availability of knowledge may differ from its use. In particular, the
growth in publishing and patenting may lead scientists and inventors
to focus on narrower slices of previous work
18,46
, thereby limiting the
‘effective’ stock of knowledge. Using three proxies, we document a
decline in the use of previous knowledge among scientists and inventors
(Fig.6). First, we see a decline in the diversity of work cited (Fig.6a,d),
indicating that contemporary science and technology are engaging
with narrower slices of existing knowledge. Moreover, this decline in
diversity is accompanied by an increase in the share of citations to the
1% most highly cited papers and patents (Fig.6a (i),d(i)), which are also
decreasing in semantic diversity (Fig.6a (ii),d (ii)). Over time, scientists
and inventors are increasingly citing the same previous work, and that
previous work is becoming more topically similar. Second, we see an
increase in self-citation (Fig.6b,e), a common proxy for the continua-
tion of one’s pre-existing research stream
47–49
, which is consistent with
scientists and inventors relying more on highly familiar knowledge.
Third, the mean age of work cited, a common measure for the use of
dated knowledge
50–52
, is increasing (Fig.6c,f), suggesting that scientists
and inventors may be struggling to keep up with the pace of knowledge
expansion and instead relying on older, familiar work. All three indica-
tors point to a consistent story: a narrower scope of existing knowledge
is informing contemporary discovery and invention.
Results from a subsequent series of regression models suggest that
use of less diverse work, more of one’s own work and older work are all
negatively associated with disruption (Methods, Extended Data Table1
and Supplementary Table3), a pattern that holds even after accounting
for the average age and number of previous works produced by team
members. When the range of work used by scientists and inventors
narrows, disruptive activity declines.
Discussion
In summary, we report a marked decline in disruptive science and
technology over time. Our analyses show that this trend is unlikely to
be driven by changes in citation practices or the quality of published
work. Rather, the decline represents a substantive shift in science and
technology, one that reinforces concerns about slowing innovative
activity. We attribute this trend in part to scientists’ and inventors’ reli-
ance on a narrower set of existing knowledge. Even though philosophers
Bednorz and Müller (1986)
Sanger et al. (1977)
Riess et al. (1998)
Kohn and Sham (1965)
–1.0
–0.8
–0.6
–0.4
–0.2
0
0.2
0.4
0.6
0.8
1.0
1900 1920 1940 1960 1980 2000
Year
CD
5
Chemistry
Medicine Physics
Dirac (1928): Discovery of the Dirac equation of relativistic quantum theory
Watson and Crick (1953): Discovery of the structure of the DNA
Kohn and Sham (1965): Development of a new method for calculating electronic structure
Sanger et al. (1977): Development of a new method for mapping the order of nucleotides
Katsuki and Sharpless (1980): Development of asymmetric epoxidation
Saiki et al. (1985): Discovery of polymerase chain reaction
Bednorz and Müller (1986): Discovery of superconductivity in cermaic materials
Riess et al. (1998): Discovery of the accelerating expansion of the universe
Dirac (1928)
Watson and Crick (1953)
Saiki et al. (1985)
Katsuki and Sharpless (1980)
Science
Nature
PNAS
0
0.25
0.50
Year
Fig. 5 | CD index of high-quality science over time. This figure shows changes
in CD
5
over time for papers published in Nature, Proceedings of the National
Academy of Sciences (PNAS) and Science (inset plot, n = 223,745) and Nobel Prize-
winning papers (main plot, n = 635), with several notable examples
31,32,58,62–66
highlighted. Colours indicate the three different journals in the inset plot;
colours indicate the three different fields in which the Nobel Prize is awarded in
the main plot. Shaded bands correspond to 95% confidence intervals. For
historical completeness, we plot CD index scores for all Nobel papers back to
1900 (the first year in which the prize was awarded); however, our main analyses
begin in the post-1945 era, when the WoS data are generally more reliable. The
figure indicates that changes in the quality of published science over time is
unlikely to be responsible for the decline in disruption.
Nature | Vol 613 | 5 January 2023 | 143
of science may be correct that the growth of knowledge is an endog-
enous process—wherein accumulated understanding promotes future
discovery and invention—engagement with a broad range of extant
knowledge is necessary for that process to play out, a requirement
that appears more difficult with time. Relying on narrower slices of
knowledge benefits individual careers
53
, but not scientific progress
more generally.
Moreover, even though the prevalence of disruptive works has
declined, we find that the sheer number has remained stable. On the
one hand, this result may suggest that there is a fixed ‘carrying capac-
ity’ for highly disruptive science and technology, in which case, policy
interventions aimed at increasing such work may prove challenging. On
the other hand, our observation of considerable churn in the underly-
ing fields responsible for producing disruptive science and technol-
ogy suggests the potential importance of factors such as the shifting
interests of funders and scientists and the ‘ripeness’ of scientific and
technologicalknowledge for breakthroughs, in which case the produc-
tion of disruptive work may be responsive to policy levers. In either
case, the stability we observe in the sheer number of disruptive papers
and patents suggests that science and technology do not appear to
have reached the end of the ‘endless frontier’. Room remains for the
regular rerouting that disruptive works contribute to scientific and
technological progress.
Our study is not without limitations. Notably, even though research
to date supports the validity of the CD index
12,34
, it is a relatively new
indicator of innovative activity and will benefit from future work on
its behaviour and properties, especially across data sources and con-
texts. Studies that systematically examine the effect of different citation
practices
54,55
, which vary across fields, would be particularly informative.
Overall, our results deepen understanding of the evolution of knowl-
edge and may guide career planning and science policy. To promote
disruptive science and technology, scholars may be encouraged to read
widely and given time to keep up with the rapidly expanding knowl-
edge frontier. Universities may forgo the focus on quantity, and more
strongly reward research quality
56
, and perhaps more fully subsidize
year-long sabbaticals. Federal agencies may invest in the riskier and
longer-term individual awards that support careers and not simply spe-
cific projects
57
, giving scholars the gift of time needed to step outside
1950 1960 1970 1980 1990 2000 2010
Year
0.94
0.95
0.96
0.97
0.98
0.99
Diversity of work cited
a
0.05
0.15
(i)
0.1
0.2
(ii)
1950 1960 1970 1980 1990 2000 2010
Year
0
1
2
3
4
5
Mean self-citations per paper
b
1950 1960 1970 1980 1990 2000
2010
Year
0
2
4
6
8
10
12
14
Mean age of work cited
c
Life sciences and biomedicine
Physical sciences
Social sciences
Technology
1980 1990 2000 2010
Year
0.94
0.95
0.96
0.97
0.98
0.99
Diversity of work cited
d
0.05
0.15
(i)
Share
of cites
to 1%
most cited
0.1
0.2
(ii)
Semantic
diversity
of 1%
most cited
1980 1990 2000 2010
Year
0
1
2
3
4
5
Mean self-citations per patent
e
1980 1990 2000 2010
Year
0
2
4
6
8
10
12
14
Mean age of work cited
f
Chemical
Computers and communications
Drugs and medical
Electrical and electronic
Mechanical
Papers
Patents
Share
of cites
to 1%
most cited
Semantic
diversity
of 1%
most cited
Fig. 6 | Papers and patents are using narrower portions of existing
knowledge. a–f, Changes in the level of diversity of existing scientific and
technological knowledge use among papers (a, n = 264 WoS research area × year
observations; b and c, n = 24,659,076 papers) and patents (d, 229 NBER
technology category × year observations; e and f, n = 3,912,353 patents) based
on following measures: diversity of work cited (a and d), mean number of
self-citations (b and e) and mean age of cited work (c and f). Shaded bands
(b,c,e and f) correspond to 95% confidence intervals. The inset plots of a and
d show changes in the share of citations to the top 1% most highly cited papers
(a(i) and d(i)) and in the semantic diversity of the top 1% most cited over time
(a(ii) and d(ii)). Values of both measures are computed within field and year,
and are subsequently averaged across fields for plotting. Semantic diversity is
based on paper and patent titles; values correspond to the ratio of the standard
deviation to the mean pairwise cosine similarity (that is, the coefficient of
variation) among the titles of the 1% most cited papers and patents by field and
year. To enable semantic comparisons, titles were vectorized using pretrained
word embeddings. For papers, lines are shown for each WoS research area; for
patents, lines are shown for each NBER technology category. In subsequent
regression analyses using these measures, we find that using less diverse work,
more of one’s own work and older work is associated with less disruptive papers
and patents (Methods and Extended Data Table1).
144 | Nature | Vol 613 | 5 January 2023
Article
the fray, inoculate themselves from the publish or perish culture, and
produce truly consequential work. Understanding the decline in dis-
ruptive science and technology more fully permits a much-needed
rethinking of strategies for organizing the production of science and
technology in the future.
Online content
Any methods, additional references, Nature Portfolio reporting summa-
ries, source data, extended data, supplementary information, acknowl-
edgements, peer review information; details of author contributions
and competing interests; and statements of data and code availability
are available at https://doi.org/10.1038/s41586-022-05543-x.
1. Fleming, L. Recombinant uncertainty in technological search. Manage. Sci. 47, 117–132
(2001).
2. Schumpeter, J. Capitalism, Socialism and Democracy (Perennial, 1942).
3. Koyré, A. An unpublished letter of Robert Hooke to Isaac Newton. ISIS 43, 312–337 (1952).
4. Popper, K. Conjectures and Refutations: The Growth of Scientiic Knowledge (Routledge,
2014).
5. Fleck, L. Genesis and Development of a Scientiic Fact (Univ. Chicago Press, 2012).
6. Acemoglu, D., Akcigit, U. & Kerr, W. R. Innovation network. Proc. Natl Acad. Sci. USA 113,
11483–11488 (2016).
7. Weitzman, M. L. Recombinant growth. Q. J. Econ. 113, 331–360 (1998).
8. Tria, F., Loreto, V., Servedio, V. D. P. & Strogatz, S. H. The dynamics of correlated novelties.
Sci. Rep. 4, 1–8 (2014).
9. Fink, T. M. A., Reeves, M., Palma, R. & Farr, R. S. Serendipity and strategy in rapid
innovation. Nat. Commun. 8, 1–9 (2017).
10. Pammolli, F., Magazzini, L. & Riccaboni, M. The productivity crisis in pharmaceutical R&D.
Nat. Rev. Drug Discov. 10, 428–438 (2011).
11. Bloom, N., Jones, C. I., Van Reenen, J. & Webb, M. Are ideas getting harder to ind? Am.
Econ. Rev. 110, 1104–1144 (2020).
12. Funk, R. J. & Owen-Smith, J. A dynamic network measure of technological change.
Manage. Sci. 63, 791–817 (2017).
13. Bornmann, L., Devarakonda, S., Tekles, A. & Chacko, G. Are disruption index indicators
convergently valid? The comparison of several indicator variants with assessments by
peers. Quant. Sci. Stud. 1, 1242–1259 (2020).
14. Uzzi, B., Mukherjee, S., Stringer, M. & Jones, B. Atypical combinations and scientiic
impact. Science 342, 468–472 (2013).
15. Leydesdorff, L., Tekles, A. & Bornmann, L. A proposal to revise the disruption index. Prof.
Inf. 30, e300121 (2021).
16. Lu, C. etal. Analyzing linguistic complexity and scientiic impact. J. Informetr. 13, 817–829
(2019).
17. Hofstra, B. etal. The diversity–innovation paradox in science. Proc. Natl Acad. Sci. USA
117, 9284–9291 (2020).
18. Jones, B. F. The burden of knowledge and the ‘death of the renaissance man’: is
innovation getting harder? Rev. Econ. Stud. 76, 283–317 (2009).
19. Gordon, R. J. The Rise and Fall of American Growth (Princeton Univ. Press, 2016).
20. Chu, J. S. G. & Evans, J. A. Slowed canonical progress in large ields of science. Proc. Natl
Acad. Sci. USA 118, e2021636118 (2021).
21. Packalen, M. & Bhattacharya, J. NIH funding and the pursuit of edge science. Proc. Natl
Acad. Sci. USA 117, 12011–12016 (2020).
22. Jaffe, A. B. & Lerner, J. Innovation and its Discontents: How Our Broken Patent System Is
Endangering Innovation and Progress, and What To Do About It (Princeton Univ. Press,
2011).
23. Horgan, J. The End of Science: Facing the Limits of Knowledge in the Twilight of the
Scientiic Age (Basic Books, 2015).
24. Collison, P. & Nielsen, M. Science Is Getting Less Bang for its Buck (Atlantic, 2018).
25. Nolan, A. Artiicial intelligence and the future of science. oecd.ai, https://oecd.ai/en/
wonk/ai-future-of-science (25 October 2021).
26. Effective Policies to Foster High-risk/High-reward Research. OECD Science, Technology,
and Industry Policy Papers (OECD, 2021).
27. Cowen, T. The Great Stagnation: How America Ate All the Low-Hanging Fruit of Modern
History, Got Sick, and Will (Eventually) Feel Better (Penguin, 2011).
28. Einstein, A. The World As I See It (Citadel Press, 1949).
29. Arthur, W. B. The structure of invention. Res. Policy 36, 274–287 (2007).
30. Tushman, M. L. & Anderson, P. Technological discontinuities and organizational
environments. Adm. Sci. Q. 31, 439–465 (1986).
31. Kohn, W. & Sham, L. J. Self-consistent equations including exchange and correlation
effects. Phys. Rev. 140, A1133 (1965).
32. Watson, J. D. & Crick, F. H. C. Molecular structure of nucleic acids: a structure for
deoxyribose nucleic acid. Nature 171, 737–738 (1953).
33. Bornmann, L. & Tekles, A. Disruption index depends on length of citation window. Prof.
Inf. 28, e280207 (2019).
34. Wu, L., Wang, D. & Evans, J. A. Large teams develop and small teams disrupt science and
technology. Nature 566, 378–382 (2019).
35. Kuhn, T. S. The Structure of Scientiic Revolutions (Univ. Chicago Press, 1962).
36. Brad Wray, K. Kuhn and the discovery of paradigms. Philos. Soc. Sci. 41, 380–397 (2011).
37. Ioannidis, J. P. A. Why most published research indings are false. PLoS Med. 2, e124
(2005).
38. Li, J., Yin, Y., Fortunato, S. & Wang, D. A dataset of publication records for Nobel laureates.
Sci. Data 6, 1–10 (2019).
39. Bornmann, L. & Marx, W. Methods for the generation of normalized citation impact scores
in bibliometrics: which method best relects the judgements of experts? J. Informetr. 9,
408–418 (2015).
40. Waltman, L. A review of the literature on citation impact indicators. J. Informetr. 10,
365–391 (2016).
41. Waltman, L. & van Eck, N. J. in Springer Handbook of Science and Technology Indicators
(eds. Glänzel, W. etal.) 281–300 (Springer, 2019).
42. Bornmann, L. How can citation impact in bibliometrics be normalized? A new approach
combining citing-side normalization and citation percentiles. Quant. Sci. Stud. 1,
1553–1569 (2020).
43. Petersen, A. M., Pan, R. K., Pammolli, F. & Fortunato, S. Methods to account for citation
inlation in research evaluation. Res. Policy 48, 1855–1865 (2019).
44. Bornmann, L. & Mutz, R. Growth rates of modern science: a bibliometric analysis based
on the number of publications and cited references. J. Assoc. Inf. Sci. Technol. 66,
2215–2222 (2015).
45. Bornmann, L., Haunschild, R. & Mutz, R. Growth rates of modern science: a latent
piecewise growth curve approach to model publication numbers from established and
new literature databases. Humanit. Soc. Sci. Commun. 8, 1–15 (2021).
46. Jones, B. F. & Weinberg, B. A. Age dynamics in scientiic creativity. Proc. Natl Acad. Sci.
USA 108, 18910–18914 (2011).
47. Bonzi, S. & Snyder, H. Motivations for citation: a comparison of self citation and citation to
others. Scientometrics 21, 245–254 (1991).
48. Fowler, J. & Aksnes, D. Does self-citation pay? Scientometrics 72, 427–437 (2007).
49. King, M. M., Bergstrom, C. T., Correll, S. J., Jacquet, J. & West, J. D. Men set their own cites
high: gender and self-citation across ields and over time. Socius 3, 2378023117738903
(2017).
50. Mukherjee, S., Romero, D. M., Jones, B. & Uzzi, B. The nearly universal link between the
age of past knowledge and tomorrow’s breakthroughs in science and technology: the
hotspot. Sci. Adv. 3, e1601315 (2017).
51. Merton, R. K. Singletons and multiples in scientiic discovery: a chapter in the sociology
of science. Proc. Am. Philos. Soc. 105, 470–486 (1961).
52. Wang, D., Song, C. & Barabási, A.-L. Quantifying long-term scientiic impact. Science 342,
127–132 (2013).
53. Leahey, E. Not by productivity alone: how visibility and specialization contribute to
academic earnings. Am. Sociol. Rev. 72, 533–561 (2007).
54. Tahamtan, I. & Bornmann, L. Core elements in the process of citing publications: conceptual
overview of the literature. J. Informetr. 12, 203–216 (2018).
55. Tahamtan, I. & Bornmann, L. What do citation counts measure? An updated review of
studies on citations in scientiic documents published between 2006 and 2018.
Scientometrics 121, 1635–1684 (2019).
56. Bhattacharya, J. & Packalen, M. Stagnation and Scientiic Incentives (Working Paper
26752), https://www.nber.org/papers/w26752 (2020).
57. Azoulay, P., Graff Zivin, J. S. & Manso, G. Incentives and creativity: evidence from the
academic life sciences. RAND J. Econ. 42, 527–554 (2011).
58. Baltimore, D. Viral RNA-dependent DNA polymerase: RNA-dependent DNA polymerase in
virions of RNA tumour viruses. Nature 226, 1209–1211 (1970).
59. Page, L. Method for node ranking in a linked database. US patent 6,285,999 (2001).
60. Axel, R., Wigler, M. H. & Silverstein, S. J. Processes for inserting DNA into eucaryotic cells
and for producing proteinaceous materials. US patent 4,634,665 (1983).
61. Hawbaker, M. S. Soybean variety SE90346. US patent 6,958,436 (2005).
62. Katsuki, T. & Sharpless, K. B. The irst practical method for asymmetric epoxidation. J. Am.
Chem. Soc. 102, 5974–5976 (1980).
63. Riess, A. G., etal. Observational evidence from supernovae for an accelerating universe
and a cosmological constant. Astron. J. 116, 1009 (1998).
64. Dirac, P. A. M. The quantum theory of the electron. Proc. R. Soc. Lond. A Math. Phys. Sci.
117, 610–624 (1928).
65. Sanger, F., Nicklen, S. & Coulson, A. R. DNA sequencing with chain-terminating inhibitors.
Proc. Natl Acad. Sci. USA 74, 5463–5467 (1977).
66. Bednorz, J. G. & Müller, K. A. Possible high T
c
superconductivity in the Ba-La-Cu-O system.
Z. Phys. B Condens. Matter 64, 189–193 (1986).
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in
published maps and institutional afiliations.
Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this
article under a publishing agreement with the author(s) or other rightsholder(s); author
self-archiving of the accepted manuscript version of this article is solely governed by the
terms of such publishing agreement and applicable law.
© The Author(s), under exclusive licence to Springer Nature Limited 2023
Methods
WoS data
We limit our focus to research papers published between 1945 and 2010.
Although the WoS data begin in the year 1900, the scale and social organ-
ization of science shifted markedly in the post-war era, thereby making
comparisons with the present difficult and potentially misleading
67–69
.
We end our analyses of papers in 2010 because some of our measures
require several subsequent years of data following paper publication.
The WoS data archive 65 million documents published in 28,968 journals
between 1900 and 2017 and 735 million citations among them. In addi-
tion, the WoS data include the titles and the full text of abstracts for 65
and 29 million records, respectively, published between 1913 and 2017.
After eliminating non-research documents (for example, book reviews
and commentaries) and subsetting the data to the 1945–2010 window,
the analytical sample consists of n = 24,659,076 papers.
Patents View data
We limit our focus to patents granted from 1976, which is the earliest
year for which machine-readable records are available in the Patents
View data. As we did with papers, we end our analyses in 2010 because
some measures require data from subsequent years for calculation.
The Patents View data are the most exhaustive source of historical
data on inventions, with information on 6.5 million patents granted
between 1976 and 2017 and their corresponding 92 million citations.
The Patents View data include the titles and abstracts for 6.5 million
patents granted between 1976 and 2017. Following previous work
12
,
we focused our attention on utility patents, which cover the vast major-
ity (91% in our data) of patented inventions. After eliminating non-utility
patents and subsetting the data to the 1976–2010 window, the analytical
sample consists of n = 3,912,353 patents.
Highly disruptive papers and patents
Observations (and claims) of slowing progress in science and technol-
ogy are increasingly common, supported not only by the evidence we
report, but also by previous research from diverse methodological and
disciplinary perspectives
10,11,18–24
. Yet as noted in the main text, there
is a tension between observations of slowing progress from aggre-
gate data on the one hand, and continuing reports of seemingly major
breakthroughs in many fields of science and technology—spanning
everything from the measurement of gravity waves to the sequencing
of the human genome—on the other. In an effort to reconcile this ten-
sion, we considered the possibility that whereas overall, discovery and
invention may be less disruptive over time, the high-level view taken in
previous work may mask considerable heterogeneity. Put differently,
aggregate evidence of slowing progress does not preclude the possibil-
ity that some subset of discoveries and inventions is highly disruptive.
To evaluate this possibility, we plot the number of disruptive papers
(Fig.4a) and patents (Fig.4b) over time, where disruptive papers and
patents are defined as those with CD
5
values >0. Within each panel, we
plot four lines, corresponding to four evenly spaced intervals—(0, 0.25],
(0.25, 0.5], (0.5, 0.75], (0.75, 1.00]—over the positive values of CD
5
. The
first two intervals therefore correspond to papers and patents that
are relatively weakly disruptive, whereas the latter two correspond to
those that are more strongly so (for example, where we may expect to
see major breakthroughs such as some of those mentioned above).
Despite major increases in the numbers of papers and patents pub-
lished each year, we see little change in the number of highly disrup-
tive papers and patents, as evidenced by the relatively flat red, green
and orange lines. Notably, this ‘conservation’ of disruptive work holds
even despite fluctuations over time in the composition of the scientific
and technological fields responsible for producing the most disrup
-
tive work (Fig.4, inset plots). Overall, these results help to account for
simultaneous observations of both major breakthroughs in many fields
of science and technology and aggregate evidence of slowing progress.
Relative contribution of field, year and authororinventor
effects
Our results show a steady decline in the disruptiveness of science and
technology over time. Moreover, the patterns we observe are generally
similar across broad fields of study, which suggests that the factors
driving the decline are not unique to specific domains of science and
technology. The decline could be driven by other factors, such as the
conditions of science and technology at a point in time or the particular
individuals who produce science and technology. For example, exog-
enous factors such as economic conditions may encourage research
or invention practices that are less disruptive. Similarly, scientists and
inventors of different generations may have different approaches,
which may result in greater or lesser tendencies for producing disrup-
tive work. We therefore sought to understand the relative contribution
of field, year and author (or inventor) factors to the decline of disruptive
science and technology.
To do so, we decomposed the relative contribution of field, year
and author fixed effects to the predictive power of regression models
of the CD index. The unit of observation in these regressions is the
author (or inventor) × year. We enter field fixed effects using granular
subfield indicators (that is, 150 WoS subject areas for papers, 138 NBER
subcategories for patents). For simplicity, we did not include additional
covariates beyond the fixed effects in our models. Field fixed effects
capture all field-specific factors that do not vary by author or year
(for example, the basic subject matter); year fixed effects capture all
year-specific factors that do not vary by field or author (for example, the
state of communication technology); author (or inventor) fixed effects
capture all author-specific factors that do not vary by field or year
(for example, the year of PhD awarding). After specifying our model,
we determine the relative contribution of field, year and author fixed
effects to the overall model adjusted R
2
using Shapley–Owen decom-
position. Specifically, given our n = 3 groups of fixed effects (field, year
and author) we evaluate the relative contribution of each set of fixed
effects by estimating the adjusted R
2
separately for the 2
n
models using
subsets of the predictors. The relative contribution of each set of fixed
effects is then computed using the Shapley value from game theory
70
.
Results of this analysis are shown in Extended Data Fig.5, for both
papers (top bar) and patents (bottom bar). Total bar size corresponds
to the value of the adjusted R
2
for the fully specified model (that is, with
all three groups of fixed effects). Consistent with our observations from
plots of the CD index over time, we observe that for both papers and
patents, field-specific factors make the lowest relative contribution to
the adjusted R
2
(0.02 and 0.01 for papers and patents, respectively).
Author fixed effects, by contrast, appear to contribute much more
to the predictive power of the model, for both papers (0.20) and pat-
ents (0.17). Researchers and inventors who entered the field in more
recent years may face a higher burden of knowledge and thus resort
to building on narrower slices of existing work (for example, because
of more specialized doctoral training), which would generally lead to
less disruptive science and technology being produced in later years,
consistent with our findings. The pattern is more complex for year
fixed effects; although year-specific factors that do not vary by field or
author hold more explanatory power than field for both papers (0.02)
and patents (0.16), they appear to be substantially more important
for the latter than the former. Taken together, these findings suggest
that relatively stable factors that vary across individual scientists and
inventors may be particularly important for understanding changes in
disruptiveness over time. The results also confirm that domain-specific
factors across fields of science and technology play a very small role in
explaining the decline in disruptiveness of papers and patents.
Alternative samples
We also considered whether the patterns we document may be artefacts
of our choice of data sources. Although we observe consistent trends
Article
in both the WoS and Patents View data, and both databases are widely
used by the Science of Science community, our results may conceivably
be driven by factors such as changes in coverage (for example, journals
added or excluded from WoS over time) or even data errors rather
than fundamental changes in science and technology. To evaluate this
possibility, we therefore calculated CD
5
for papers in four additional
databases—JSTOR, the American Physical Society corpus, Microsoft
Academic Graph and PubMed. We included all records from 1930 to
2010 from PubMed (16,774,282 papers), JSTOR (1,703,353 papers)
and American Physical Society (478,373 papers). The JSTOR data were
obtained via a special request from ITHAKA, the data maintainer (http://
www.ithaka.org), as were the American Physical Society data (https://
journals.aps.org/datasets). We downloaded the Microsoft Academic
Graph data from CADRE at Indiana University (https://cadre.iu.edu/).
The PubMed data were downloaded from the National Library of Medi-
cine FTP server (ftp://ftp.ncbi.nlm.nih.gov/pubmed/baseline). Owing
to the exceptionally large scale of Microsoft Academic Graph and the
associated computational burden, we randomly extracted 1 million
papers. As shown in Extended Data Fig.6, the downward trend in dis-
ruptiveness is evident across all samples.
Alternative bibliometric measures
Several recent papers have introduced alternative specifications of
theCD index
12
. We evaluated whether the declines in disruptiveness
we observe are corroborated using two alternative variations. One
criticism of the CD index has been that the number of papers that cite
only the focal paper’s references dominates the measure
13
. Bornmann
etal.
13
proposes
as a variant that is less susceptible to this issue.
Another potential weakness of the CD index is that it could be very
sensitive to small changes in the forward citation patterns of papers
that make no backward citations
15
. Leydesdorff etal.
15
suggests DI* as
an alternate indicator of disruption that addresses this issue. Therefore,
we calculated
where l = 5 and DI* for 100,000 randomly drawn
papers and patents each from our analytic sample. Results are pre-
sented in Extended Data Fig.7a (papers) and b (patents). The blue lines
indicate disruption based on Bornmann etal.
13
and the orange lines
indicate disruption based on Leydesdorff etal.
15
. Across science and
technology, the two alternative measures both show declines in disrup-
tion over time, similar to the patterns observed with the CD index.
Taken together, these results suggest that the declines in disruption
we document are not an artefact of our particular operationalization.
Robustness to changes in publication, citation and authorship
practices
We also considered whether our results may be attributable to changes
in publication, citation or authorship practices, rather than by sub-
stantive shifts in discovery and invention. Perhaps most critically, as
noted in the main text, there has been a marked expansion in publish-
ing and patenting over the period of our study. This expansion has
naturally increased the amount of previous work that is relevant to
current science and technology and therefore at risk of being cited,
a pattern reflected in the marked increase in the average number of
citations made by papers and patents (that is, papers and patents are
citing more previous work than in previous eras)
44,45
. Recall that the
CD index quantifies the degree to which future work cites a focal work
together with its predecessors (that is, the references in the bibliography
of the focal work). Greater citation of a focal work independently of its
predecessors is taken to be evidence of a social process of disruption.
As papers and patents cite more previous work, however, the prob-
ability of a focal work being cited independently of its predecessors
may decline mechanically; the more citations a focal work makes, the
more likely future work is to cite it together with one of its predeces
-
sors, even by chance. Consequently, increases in the number of papers
and patents available for citing and in the average number of citations
made by scientists and inventors may contribute to the declining values
of the CD index. In short, given the marked changes in science and
technology over our long study window, the CD index of papers and
patents published in earlier periods may not be directly comparable
to those of more recent vintage, which could in turn render our conclu-
sions about the decline in disruptive science and technology suspect.
We addressed these concerns using three distinctive but complemen-
tary approaches—normalization, regression adjustment and simulation.
Veriication using normalization. First, following common practice
in bibliometric research
39–43
, we developed two normalized versions
of the CD index, with the goal of facilitating comparisons across time.
Among the various components of the CD index, we focused our
attention on the count of papers or patents that only cite the focal
work’s references (N
k
), as this term would seem most likely to scale with
the increases in publishing and patenting and in the average number of
citations made by papers and patents to previous work
13
. Larger values
of N
k
lead to smaller values of the CD index. Consequently, marked
increases in N
k
over time, particularly relative to other components
of the measure, may lead to a downward bias, thereby inhibiting our
ability to accurately compare disruptive science and technology in
later years with earlier periods.
Our two normalized versions of the CD index aim to address this
potential bias by attenuating the effect of increases in N
k
. In the first
version, which we call ‘Paper normalized’, we subtract from N
k
the num-
ber of citations made by the focal paper or patent to previous work
(N
b
). The intuition behind this adjustment is that when a focal paper
or patent cites more previous work, N
k
is likely to be larger because
there are more opportunities for future work to cite the focal paper or
patent’s predecessors. This increase in N
k
would result in lower values
of the CD index, although not necessarily as a result of the focal paper
or patent being less disruptive. In the second version, which we call
‘field × year normalized’, we subtract N
k
by the average number of back-
ward citations made by papers or patents in the focal paper or patent’s
WoS research area or NBER technology category, respectively, during
its year of publication (we label this quantity
N
b
mean
). The intuition
behind this adjustment is that in fields and time periods in which there
is a greater tendency for scientists and inventors to cite previous work,
N
k
is also likely to be larger, thereby leading to lower values of the CD
index, although again not necessarily as a result of the focal paper or
patent being less disruptive. In cases in which either N
b
or
N
b
mean
exceed
the value of N
k
, we set N
k
to 0 (that is, N
k
is never negative in the normal-
ized measures). Both adaptations of the CD index are inspired by estab-
lished approaches in the scientometrics literature, and may be
understood as a form of ‘citing side normalization’ (that is, normaliza-
tion by correcting for the effect of differences in lengths of references
lists)
40
.
In Extended Data Fig.8, we plot the average values of both normalized
versions of the CD index over time, separately for papers (Extended
Data Fig.8a) and patents (Extended Data Fig.8d). Consistent with our
findings reported in the main text, we continue to observe a decline
in the CD index over time, suggesting that the patterns we observe in
disruptive science and technology are unlikelyto be driven bychanges
in citation practices.
Veriication using regression adjustment. Second, we adjusted
for potential confounding using a regression-based approach. This
approach complements the bibliometric normalizations just described
by allowing us to account for a broader array of changes in publica-
tion, citation and authorship practices in general (the latter of which
is not directly accounted for in either the normalization approach or
the simulation approach described next), and increases the amount
of previous work that is relevant to current science and technol-
ogy in particular. In Supplementary Table1, we report the results of
regression models predicting CD
5
for papers (Models 1–4) and patents
(Models 5–8), with indicator variables included for each year of our
study window (the reference categories are 1945 and 1980 for papers
and patents, respectively). Models 1 and 4 are the baseline models, and
include no other adjustments beyond the year indicators. In Models 2
and 5, we add subfield fixed effects (WoS subject areas for papersand
NBER technology subcategories for patents). Finally, in Models 3–4 and
7–8, we add control variables for several field × year level—number of
new papersorpatents, mean number of papersorpatents cited, mean
number of authorsorinventors per paper—and paper-orpatent-level—
number of papersorpatents cited—characteristics, thereby enabling
more robust comparisons in patterns of disruptive science and tech-
nology over the long time period spanned by our study. For the paper
models, we also include a paper-level control for the number of unlinked
references (that is, the number of citations to works that are not indexed
in WoS). We find that the inclusion of these controls improves model
fit, as indicated by statistically significant Wald tests presented below
the relevant models.
Across all eight models shown in Supplementary Table1, we find that
the coefficients on the year indicators are statistically significant and
negative, and growing in magnitude over time, which is consistent with
the patterns we reported based on unadjusted CD
5
values index in the
main text (Fig.2). In Extended Data Fig.8, we visualize the results of
our regression-based approach by plotting the predicted CD
5
values
separately for each of the year indicators included in Models 4 (papers)
and 8 (patents). To enable comparisons with raw CD
5
values shown in
the main text, we present the separate predictions made for each year
as a line graph. As shown in the figure, we continue to observe declining
values of the CD index across papers and patents, even when accounting
for changes in publication, citation and authorship practices.
Veriication using simulation. Third, following related work in the
Science of Science
14,71–73
, we considered whether our results may be an
artefact of changing patterns in publishing and citation practices by
using a simulation approach. In essence, the CD index measures disrup-
tion by characterizing the network of citations around a focal paper or
patent. However, many complex networks, even those resulting from
random processes, exhibit structures that yield non-trivial values on
common network measures (for example, clustering)
74–76
. During the
period spanned by our study, the citation networks of science and
technology experienced significant change, with marked increases in
both the numbers of nodes (that is, papers or patents) and edges (that
is, citations). Thus, rather than reflecting a meaningful social process,
the observed declines in disruption may result from these structural
changes in the underlying citation networks.
To evaluate this possibility, we followed standard techniques from
network science
75,77
and conducted an analysis in which we recomputed
the CD index on randomly rewired citation networks. If the patterns
we observe in the CD index are the result of structural changes in the
citation networks of science and technology (for example, growth in
the number of nodes or edges) rather than a meaningful social process,
then these patterns should also be visible in comparable random net-
works that experience similar structural changes. Therefore, finding
that the patterns we see in the CD index differ for the observed and
random citation networks would serve as evidence that the decline in
disruption is not an artefact of the data.
We began by creating copies of the underlying citation network on
which the values of the CD index used in all analyses reported in the main
text were based, separately for papers and patents. For each citation
network (one for papers, one for patents), we then rewired citations
using a degree-preserving randomization algorithm. In each iteration of
the algorithm, two edges (for example, A–B and C–D) are selected from
the underlying citation network, after which the algorithm attempts to
swap the two endpoints of the edges (for example, A–B becomes A–D,
and C–D becomes C–B). If the degree centrality of A, B, C and D remains
the same after the swap, the swap is retained; otherwise, the algorithm
discards the swap and moves on to the next iteration. When evaluating
degree centrality, we consider ‘in-degree’ (that is, citations from other
papersorpatents to the focal paperorpatent) and ‘out-degree’ (that
is, citations from the focal paperorpatent to other papersorpatents)
separately. Furthermore, we also required that the age distribution
of citing and cited papersorpatents was identical in the original and
rewired networks. Specifically, swaps were only retained when the
publication year of the original and candidate citations was the same.
In light of these design choices, our rewiring algorithm should be seen
as fairly conservative, as it preserves substantial structure from the
original network. There is no scholarly consensus on the number of
swaps necessary to ensure the original and rewired networks are suf-
ficiently different from one another; the rule we adopt here is 100 × m,
where m is the number of edges in the network being rewired.
Following previous work
14
, we created ten rewired copies of the
observed citation networks for both papers and patents. After creat-
ing these rewired citation networks, we then recomputed CD
5
. Owing
to the large scale of the WoS data, we base our analyses on a random
subsample of ten million papers; CD
5
was computed on the rewired
network for all patents. For each paper and patent, we then compute
a z score that compares the observed CD
5
value to those of the same
paper or patent in the ten rewired citation networks. Positive z scores
indicate that the observed CD
5
value is greater (that is, more disruptive)
than would be expected by chance; negative z scores indicate that the
observed values are lesser (that is, more consolidating).
The results of these analyses are shown in Extended Data Fig.8, sepa-
rately for papers (Extended Data Fig.8c) and patents (Extended Data
Fig.8f). Lines correspond to the average z score among papers or pat-
ents published in the focal year. The plots reveal a pattern of change in
the CD index over and beyond that ‘baked in’ to the changing structure
of the network. We find that on average, papers and patents tend to
be less disruptive than would be expected by chance, and moreover,
the gap between the observed CD index values and those from the
randomly rewired networks is increasing over time, which is consistent
with our findings of a decline in disruptive science and technology.
Taken together, the results of the foregoing analyses suggest that
although there have been marked changes in science and technology
over the course of our long study window, particularly with respect to
publication, citation and authorship practices, the decline in disrup-
tive science and technology that we document using the CD index is
unlikely to be an artefact of these changes, and instead represents a
substantive shift in the nature of discovery and invention.
Regression analysis
We evaluate the relationship between disruptiveness and the use of pre-
vious knowledge using regression models, predicting CD
5
for individual
papers and patents, based on three indicators of previous knowledge
use—the diversity of work cited, mean number of self-citations and
mean age of work cited. Our measure of the diversity of work cited is
measured at the field × year level; all other variables included in the
regressions are defined at the level of the paper or patent. To account
for potential confounding factors, our models included year and field
fixed effects. Year fixed effects account for time variant factors that
affect all observations (papers or patents) equally (for example, global
economic trends). Field fixed effects account for field-specific factors
that do not change over time (for example, some fields may intrinsi-
cally value disruptive work over consolidating ones). In contrast to our
descriptive plots, for our regression models, we adjust for field effects
using the more granular 150 WoS ‘extended subjects’ (for example,
‘biochemistry and molecular biology’, ‘biophysics’, ‘biotechnology and
applied microbiology’, ‘cell biology’, ‘developmental biology’, ‘evolu-
tionary biology’ and ‘microbiology’ are extended subjects within the
life sciences and biomedicine research area) and 38 NBER technology
subcategories (for example, ‘agriculture’, ‘food’, ‘textile’; ‘coating’; ‘gas’;
‘organic’; and ‘resins’ are subcategories within the chemistry technol-
ogy category).
Article
In addition, we also include controls for the ‘mean age of team mem-
bers’ (that is, ‘career age’, defined as the difference between the publi-
cation year of the focal paper or patent and the first year in which each
author or inventor published a paper or patent) and the ‘mean number
of previous works produced by team members’. Although increases
in rates of self-citations may indicate that scientists and inventors
are becoming more narrowly focused on their own work, these rates
may also be driven in part by the amount of previous work available
for self-citing. Similarly, although increases in the age of work cited
in papers and patents may indicate that scientists and inventors are
struggling to keep up, they may also be driven by the rapidly aging
workforce in science and technology
78,79
. For example, older scientists
and inventors may be more familiar with or more attentive to older
work, or may actively resist change
80
. These control variables help to
account for these alternative explanations.
Supplementary Table3 shows summary statistics for variables used in
the ordinary-least-squares regression models. The diversity of work cited
is measured by normalized entropy, which ranges from 0 to 1. Greater
values on this measure indicate a more uniform distribution of citations
to a wider range of existing work; lower values indicate a more concen-
trated distribution of citations to a smaller range of existing work. The
tables show that the normalized entropy in a given field and year has a
nearly maximal average entropy of 0.98 for both science and technol-
ogy. About 16% of papers cited in a paper are by an author of the focal
paper; the corresponding number for patents is about 7%. Papers tend
to rely on older work and work that varies more greatly in age (measured
by standard deviation) than patents. In addition, the average CD
5
of a
paper is 0.04 whereas the average CD
5
of a patent is 0.12, meaning that
the average paper tends to be less disruptive than the average patent.
We find that using more diverse work, less of one’s own work and
older work tends to be associated with the production of more disrup-
tive science and technology, even after accounting for the average age
and number of previous works produced by team members. These
findings are based on our regression results, shown in Extended Data
Table1. Models 6 and 12 present the full regression models. The models
indicate a consistent pattern for both science and technology, wherein
the coefficients for diversity of work cited are positive and significant
for papers (0.159, P < 0.01) and patents (0.069, P < 0.01), indicating that
in fields in which there is more use of diverse work, there is greater
disruption. Holding all other variables at their means, thepredicted
CD
5
of papers and patents increases by 303.5% and 1.3%, respectively,
when the diversity of work cited increases by 1 s.d. The coefficients of
the ratio of self-citations to total work cited is negative and significant
for papers (−0.011, P < 0.01) and patents (−0.060, P < 0.01), showing that
when researchers or inventors rely more on their own work, discovery
and invention tends to be less disruptive. Again holding all other vari-
ables at their means, the predicted CD
5
of papers and patents decreases
by 622.9% and 18.5%, respectively, with a 1 s.d. increase in the ratio. The
coefficients of the interaction between mean age of work cited and
dispersion in age of work cited is positive and significant for papers
(0.000, P < 0.01) and patents (0.001, P < 0.01), suggesting that—holding
the dispersion of the age of work cited constant—papers and patents
that engage with older work are more likely to be disruptive. The pre-
dicted CD
5
of papers and patents increases by a striking 2,072.4% and
58.4%, respectively, when the mean age of work cited increases by 1 s.d.
(about nine and eight years for papers and patents, respectively), again
holding all other variables at their means. In summary, the regression
results suggest that changes in the use of previous knowledge may
contribute to the production of less disruptive science and technology.
Reporting summary
Further information on research design is available in theNature Port-
folio Reporting Summary linked to this article.
Data availability
Data associated with this study are freely available in a public reposi-
tory at https://doi.org/10.5281/zenodo.7258379. Our study draws on
data from six sources: the American Physical Society, JSTOR, Microsoft
Academic Graph, Patents View, PubMed and WoS. Data from Microsoft
Academic Graph, Patents View and PubMed are publicly available,
and our repository includes complete data for analyses from these
sources. Data from the American Physical Society, JSTOR and WoS are
not publicly available, and were used under licence from their respec-
tive publishers. To facilitate replication, our repository includes limited
versions of the data from these sources, which will enable calculation of
basic descriptive statistics. The authors will make full versions of these
data available upon request and with permission from their respective
publishers.Source data are provided with this paper.
Code availability
Open-source code related to this study is available at https://doi.
org/10.5281/zenodo.7258379 and http://www.cdindex.info. We used
Python v.3.10.6 (pandas v.1.4.3, numpy v.1.23.1, matplotlib v.3.5.2,
seaborn v.0.11.2, spacy v.2.2, jupyterlab v.3.4.4) to wrangle, analyse
and visualize data and to conduct statistical analyses. We used MariaDB
v.10.6.4 to wrangle data. We used R v.4.2.1 (ggplot2 v.3.36, ggrepel
v.0.9.0) to visualize data. We used StataMP v.17.0 (reghdfe v.5.7.3) to
conduct statistical analyses.
67. Wuchty, S., Jones, B. F. & Uzzi, B. The increasing dominance of teams in production of
knowledge. Science 316, 1036–1039 (2007).
68. Guimera, R., Uzzi, B., Spiro, J. & Amaral, L. A. N. Team assembly mechanisms determine
collaboration network structure and team performance. Science 308, 697–702 (2005).
69. Jones, B. F., Wuchty, S. & Uzzi, B. Multi-university research teams: shifting impact,
geography, and stratiication in science. Science 322, 1259–1262 (2008).
70. Grömping, U. Estimators of relative importance in linear regression based on variance
decomposition. Am. Stat. 61, 139–147 (2007).
71. Mukherjee, S., Uzzi, B., Jones, B. & Stringer, M. A new method for identifying recombinations
of existing knowledge associated with high-impact innovation. J. Prod. Innov. Manage.
33, 224–236 (2016).
72. Christianson, N. H., Sizemore Blevins, A. & Bassett, D. S. Architecture and evolution of
semantic networks in mathematics texts. Proc. R. Soc. A 476, 20190741 (2020).
73. Newman, M. E. J. The structure of scientiic collaboration networks. Proc. Natl Acad. Sci.
USA 98, 404–409 (2001).
74. Newman, M. E. J. Scientiic collaboration networks. I. Network construction and
fundamental results. Phys. Rev. E 64, 016131 (2001).
75. Uzzi, B. & Spiro, J. Collaboration and creativity: the small world problem. Am. J. Sociol.
111, 447–504 (2005).
76. Funk, R. J. Making the most of where you are: geography, networks, and innovation in
organizations. Acad. Manage. J. 57, 193–222 (2014).
77. Barabási, A.-L. Network Science (Cambridge Univ. Press, 2016).
78. Blau, D. M. & Weinberg, B. A. Why the US science and engineering workforce is aging
rapidly. Proc. Natl Acad. Sci. USA 114, 3879–3884 (2017).
79. Cui, H., Wu, L. & Evans, J. A. Aging scientists and slowed advance. Preprint at https://doi.
org/10.48550/arXiv.2202.04044 (2022).
80. Azoulay, P., Fons-Rosen, C. & Graff Zivin, J. S. Does science advance one funeral at a time?
Am. Econ. Rev. 109, 2889–2920 (2019).
Acknowledgements This study was supported by the National Science Foundation (grant Nos.
1829168, 1932596 and 1829302).
Author contributions R.J.F. and E.L. collaboratively contributed to the conception and design
of the study. R.J.F. and M.P. collaboratively contributed to the acquisition, analysis and
interpretation of the data. R.J.F. created software used in the study. R.J.F., E.L. and M.P.
collaboratively drafted and revised the manuscript.
Competing interests The authors declare no competing interests.
Additional information
Supplementary information The online version contains supplementary material available at
https://doi.org/10.1038/s41586-022-05543-x.
Correspondence and requests for materials should be addressed to Russell J. Funk.
Peer review information Nature thanks Diana Hicks and the other, anonymous, reviewer(s) for
their contribution to the peer review of this work. Peer reviewer reports are available.
Reprints and permissions information is available at http://www.nature.com/reprints.
Extended Data Fig. 1 | Distribution of CD
5
. This figure gives an overview of the
distribution of CD
5
for papers (n = 24,659,076) and patents (n = 3,912,353).
Panels a and c show counts of papers and patents over discrete intervals of CD
5
.
Panels b and d show the distribution of CD
5
over time, within 10 (papers) and
5 (patents) year intervals, using letter-value plots. These plots are similar to
boxplots, but generally provide more reliable summaries for large datasets.
They are drawn by identifying the median of the underlying distribution and
then recursively drawing boxes outward from there in either direction that
encompass half of the remaining data.
Article
Extended Data Fig. 2 | CD index measured using alternative forward
citation windows. This figure evaluates the sensitivity of our results to the use
of different forward citation windows when computing the CD index for papers
(n = 24,659,076) and patents (n = 3,912,353). In the main text, the index is
computed based on citations made to papers and patents and their backward
references as of 5 years after the year of publication. a and c plot the CD index
using a longer, 10 year forward window, for papers and patents, respectively.
b and d plot the CD index using all forward citations made to sample papers and
patents as of the year 2017. Shaded bands correspond to 95% confidence
intervals. Overall, the results mirror those reported in the main text, although
the decline is somewhat steeper using longer forward citation windows,
suggesting our primary results may represent a more conservative estimate.
Extended Data Fig. 3 | Diversity of language use in science and technology
over time. This figure shows changes in the ratio of unique to total words (also
known as the type-token ratio) over time based on data from the abstracts of
papers (a, n = 76 WoS research area × year observations) and patents (b, n = 229
NBER technology category × year observations). For papers, lines correspond
to WoS research areas; for patents, lines correspond to NBER technology
categories. For paper abstracts, lines begin in 1992 because WoS does not
reliably record abstracts for papers published prior to the early 1990s. The ratio
of unique to total words is computed separately by field (i.e., the uniqueness of
words and total word counts are determined within WoS research areas and
NBER technology categories). If disruption is decreasing, we may plausibly
expect to see a decrease in the diversity of words used by scientists and
inventors, as discoveries and inventions will be less likely to create departures
from the status quo, and will therefore be less likely to need to introduce new
terminology. For both papers and patents, we observe declining diversity in
word use over time, which is consistent with this expectation and corroborates
our findings using the CD index.
Article
Extended Data Fig. 4 | Declining combinatorial novelty. This figure shows
changing patterns in the combinatorial novelty/conventionality of papers
(a, n = 24,659,076) and patents (b, n = 3,912,353), using a previously proposed
measure of “atypical combinations”
14
. The measure quantifies the degree to
which the prior work cited by a paper or patent would be expected by chance.
For papers, we follow prior work
14
and consider combinations of cited journals.
If a paper made three citations to prior work, and that work was published in
three different journals—Nature, Cell, and Science—then there are three
combinations—Nature × Cell, Nature × Science, and Science × Cell. To determine
the degree to which each combination would be expected by chance, the
frequency of observed pairings is compared to those in 10 “rewired” copies of
the overall citation network, using a z-score. For patents, there is no natural
analogue to journals, and therefore we consider pairings of primary United
States Patent Classification (USPC) system codes. We present the results of this
analysis following the approach of prior work
14
, which plots the cumulative
distribution function of the measure. In general, there is a rightward shift in the
cumulative distributions over time, suggesting that for both papers and
patents, combinations are more conventional than would be expected by
chance, consistent with what we would anticipate based on our results using
the CD index. For patents, there is also a smaller shift in the opposite direction
on the left side of the distribution, suggesting that novel patents in recent
decades are somewhat more novel than novel patents in earlier decades.
Overall, however, the bulk of the distribution is moving rightward, indicating
greater conventionality.
Extended Data Fig. 5 | Contribution of field, year, and author effects. This
figure shows the relative contribution of field, year, and author fixed effects
to the adjusted R
2
in regression models predicting CD
5
. The top bar shows the
results for papers (n = 80,607,091 paper × author observations); the bottom bar
shows the results for patents (n = 8,319,826 patent × inventor observations).
The results suggest that for both papers and patents, stable characteristics of
authors contribute significantly to patterns of disruptiveness. Moreover,
relatively little of the variation is accounted for by field-specific factors.
Article
Extended Data Fig. 6 | CD index over time across data sources. This figure
shows changes in CD
5
over time across four additional data sources (the WoS
[n = 24,659,076] and Patents View [n = 3,912,353] lines are included for reference):
JSTOR (n = 1,703,353), the American Physical Society corpus (n = 478,373),
Microsoft Academic Graph (n = 1,000,000), and PubMed (n = 16,774,282).
Colours indicate the six different data sources. Shaded bands correspond to
95% confidence intervals. The figure indicates that the decline in disruption is
unlikely to be driven by our sample choice of WoS papers and Patents View
patents.
Extended Data Fig. 7 | Alternative measures of disruption. This figure
shows the decline in the disruption of papers (a, n = 100,000) and patents
(b, n = 100,000) based on two alternative measures of disruption. The blue
lines calculate disruption using a measure proposed in Bornmann etal.
13
,
D
I
l
nok
where l = 5, which makes the measure more resilient to marginal changes in the
number of papers or patents that only cite the focal work’s references. The
orange lines calculate disruption using a measure proposed in Leydesdorff etal.
15
,
DI*, which makes the measure less sensitive to small changes in the forward
citation patterns of papers or patents that make no backward citations. Shaded
bands correspond to 95% confidence intervals. With both alternative
measures, we observe decreases in disruption for papers and patents,
suggesting that the decline is not an artefact of our operationalization of
disruption.
Article
Extended Data Fig. 8 | Robustness to changes in publication, citation, and
authorship practices. This figure evaluates whether declines in disruptiveness
may be attributable to changes in publication, citation, and authorship practices
for papers (n = 24,659,076) and patents (n = 3,912,353). Panels a and d adjust for
these changes using a normalization approach. We present two alternative
versions of the CD index, both of which account for the tendency for papers and
patents to cite more prior work over time. Blue lines indicate normalization at
the paper level (accounting for the number of citations made by the focal
paper/patent). Orange lines indicate normalization at the field and year level
(accounting for the mean number of citations made by papers/patents in the
focal field and year). Panels b (papers) and e (patents) adjust for changes in
publication, citation, and authorship practices using a regression approach.
The panels show predicted values of CD
5
based on regressions reported in
Models 4 (papers) and 8 (patents) of Supplementary Table1, which adjust for
field × year—Number of new papers/patents, Mean number of papers/patents
cited, Mean number of authors/inventors per paper/patent—and paper/patent-
level—Number of papers/patents cited, Number of unlinked references—
characteristics. Predictions are made separately for each year indicator
included in the models; we then connect these separate predictions with lines to
aid interpretation. Finally, Panels c (papers) and f (patents) adjust for changes
in publication, citation, and authorship practices using a simulation approach.
The panels plot z-scores that compare values of CD
5
obtained from the observed
citation networks to those obtained from randomly rewired copies of the
observed networks. Across all six panels, shaded bands correspond to 95%
confidence intervals.
Extended Data Fig. 9 | Growth of scientific and technological knowledge. This figure shows the number of papers (n = 24,659,076) published (a) and patents
(n = 3,912,353) granted (b) over time. For papers, lines correspond to WoS research areas; for patents, lines correspond to NBER technology categories.
Article
Extended Data Table 1 | Regression models of disruptiveness and the use of prior knowledge
Notes: This table evaluates the relationship between different measures of the use of prior scientiic and technological knowledge and CD
5
. Estimates are from ordinary-least-squares
regressions. Each coeficient is tested against the null hypothesis of being equal to 0 using a two-sided t-test. We do not adjust for multiple hypothesis testing. Robust standard errors
are shown in parentheses.
*p<0.1; **p<0.05; ***p<0.01.

Discussion

Tap the paper where you want to comment
In 1675, Isaac Newton remarked in a letter to rival Robert Hooke: “What DesCartes did was a good step. You have added much several ways, & especially in taking the colours of thin plates into philosophical consideration. If I have seen further it is by standing on the shoulders of Giants.” For more on the expression “standing on the shoulders of Giants”: https://en.wikipedia.org/wiki/Standing_on_the_shoulders_of_giants This paper made quite a splash when it came out. This Nature editorial contextualizes it and distills some of the response from the scientific community to its findings. Source: https://www.nature.com/articles/d41586-023-00183-1 >>For this Editorial, *Nature* spoke to a number of scholars who study science and innovation. The paper by Park and his colleagues, they say, builds on a pattern identified elsewhere in the specialist literature, and some are worried by the findings’ implications. Science and innovation are drivers of both growth and productivity, and declining disruptiveness could be linked to the sluggish productivity and economic growth being seen in many parts of the world. >>Others argue that a decline in the fraction of disruptive science shouldn’t cause concern if the absolute number of disruptive studies has remained relatively constant over time. If a greater proportion of publications are consolidating, that could just reflect the current scientific situation: in many disciplines, the fundamentals are agreed on, so most further advances will be incremental, rather than disruptive….. >>Another reason why the study by Park and his colleagues has created such resonance is that it plays into wider concerns about how science is organized. One of these is whether the division of science into ever-narrower units of knowledge is detrimental to the discovery of new paths. Critics also point to publication incentives and metrics-driven research evaluations, which steer scientific study away from risk-taking as funders, researchers and institutions take the safe option to keep the grant–publication–citation wheel turning. >> “We find that papers and patents are increasingly less likely to break with the past in ways that push science and technology in new directions. This pattern holds universally across fields and is robust across multiple different citation- and text-based metrics. Subsequently, we link this decline in disruptiveness to a narrowing in the use of previous knowledge, allowing us to reconcile the patterns we observe with the ‘shoulders of giants’ view. We find that the observed declines are unlikely to be driven by changes in the quality of published science, citation practices or field-specific factors. Overall, our results suggest that slowing rates of disruption may reflect a fundamental shift in the nature of science and technology." >>"Yet much remains unknown, not merely about the causes of slowing innovative activity, but also the depth and breadth of the phenomenon. The decline is difficult to reconcile with centuries of observation by philosophers of science, who characterize the growth of knowledge as an endogenous process, wherein previous knowledge enables future discovery, a view captured famously in Newton’s observation that if he had seen further, it was by ‘standing on the shoulders of giants." The analyses and methods shared in this paper hinge on the CD metric, so it is worth understanding its advantages, limitations, and history. >> "We quantify this distinction using a measure—the CD index—that characterizes the consolidating or disruptive nature of science and technology (Fig.1). The intuition is that if a paper or patent is disruptive, the subsequent work that cites it is less likely to also cite its predecessors; for future researchers, the ideas that went into its production are less relevant (for example, Pauling’s triple helix). If a paper or patent is consolidating, subsequent work that cites it is also more likely to cite its predecessors; for future researchers, the knowledge upon which the work builds is still (and perhaps more) relevant (for example, the theorems Kohn and Sham used). The CD index ranges from −1 (consolidating) to 1 (disruptive)." Even if the rates of disruptive work are going down, the absolute count of disruptive papers has remained relatively flat over the period studied, with the exception of life sciences, which has seen an explosion in the quantity of disruptive papers. Interesting to see that Watson and Crick's DNA discovery isn't even at the top of the CD scale! >> "Overall, our results deepen understanding of the evolution of knowledge and may guide career planning and science policy. To promote disruptive science and technology, scholars may be encouraged to read widely and given time to keep up with the rapidly expanding knowledge frontier. Universities may forgo the focus on quantity, and more strongly reward research quality, and perhaps more fully subsidize year-long sabbaticals. Federal agencies may invest in the riskier and longer-term individual awards that support careers and not simply specific projects, giving scholars the gift of time needed to step outside the fray, inoculate themselves from the publish or perish culture, and produce truly consequential work. Understanding the decline in disruptive science and technology more fully permits a much-needed rethinking of strategies for organizing the production of science and technology in the future."