 |
| String Theory |
A team of Chinese and American researchers published on arXiv an article in September 2024 summarizing their study of the development of science and technology over more than two centuries. The study analyzes 213 million scientific articles published between 1800 and 2020, along with 7.6 million patents granted between 1976 and 2020.
The result of this study is the following: while the number of scientific publications has grown exponentially, the knowledge obtained by humanity grows linearly; in other words, the speed of knowledge acquisition is constant and does not grow in the same proportion as the number of publications.
The number of
publications is easily calculable (we just have to count them), but the
knowledge acquired is more difficult to quantify. The authors of the paper have
designed what they call the Knowledge Quantification Index (KQI), a measure
related to information entropy, which measures disorder and is obtained from a
network of citations built from the millions of articles analyzed. The authors assert
that this measure (KQI) assembles numerous scientific impact metrics into one
and gains interpretability and resistance to manipulation. They
also say that its usefulness is demonstrated
for mining influential classics and laureates that are omitted by traditional metrics.
According to the
authors, the evolution of their measure (KQI) over time shows that the increase
in knowledge over the last 200 years has been and continues to be linear, and
since the number of articles published has grown exponentially in recent
decades, it can be deduced that scientific
productivity (that is, the increase in scientific knowledge for
each article published) is declining
exponentially.
As for the number
of patents, which is increasing more slowly than the number of published
articles, the use of KQI also detects linear growth, meaning that the results
obtained apply not just to science, but also to technology.
The analysis has
detected inflection points where the slope of the knowledge line changes
abruptly. These points indicate major discoveries that open up new fields of
research, such as relativity, quantum mechanics or the discovery of the
structure of DNA. When this happens, the rate of knowledge growth increases for
some time, although it remains linear, giving the impression of an exponential
increase, which in fact does not take place.
In a parallel
study whose results were published in a Nature article
in January 2023, Max Kozlov detected a similar phenomenon: what he calls disruptive science (discoveries such as those
mentioned in the previous paragraph, which change the direction of science) is
in decline, and no one knows why. The subtitle of his article reads as follows:
The proportion of publications that send a
field in a new direction has plummeted over the past half-century.
This result is in
line with what I have said in the first paragraphs of this article. It is also
in line with several previous posts in this blog, where I mentioned my feeling that scientific
productivity is declining, although my assessment was rather qualitative, and
the numerical data justifying it could be questionable, since they were not enough
in number.
Although it seems that the scientific knowledge of humanity continues to increase linearly, the fact that scientific productivity is plummeting has deplorable consequences, which may threaten the development of future science. Why? Because the scientific knowledge of each of us increases by reading publications, and if the productivity of each of them tends to zero, either we will have to read all the time to keep up to date, which means we will not be able to do research, or we will have to stop reading what others do in order to be able to do something ourselves, and in both cases the progress of science cannot help but suffer.
Thematic Thread about Science in General: Previous Next
Manuel Alfonseca
No comments:
Post a Comment