08-20-2026, 04:01 PM
The Reasonable Ineffectiveness of Mathematics
Author: Derek Abbott
Published: October 2013
Journal:Proceedings of the IEEE, Vol. 101, No. 10, pp. 2147–2153
DOI: 10.1109/JPROC.2013.2274907
Derek Abbott challenges Eugene Wigner’s famous claim about the “unreasonable effectiveness of mathematics” in describing nature. Rather than regarding the success of mathematics in physics as something mysterious or as evidence that mathematical objects have an independent Platonic existence, Abbott argues that mathematics is largely a human invention constructed to represent patterns and regularities that humans happen to observe. We naturally notice problems that our mathematical tools can solve, develop new mathematics when existing tools fail, and tend to remember successful mathematical models while overlooking the enormous number of unsuccessful ones. Thus, what appears to be a miraculous correspondence between mathematics and nature may partly be a form of selection bias.
Abbott develops Richard Hamming's earlier observations: we see what we are equipped to look for; we choose the mathematics appropriate to particular problems; science successfully addresses only a comparatively small subset of conceivable questions; and human cognition itself was shaped by evolution to understand phenomena occurring on approximately human spatial and temporal scales. Abbott adds two ideas of his own. First, mathematical laws function as a kind of lossy compression of reality: equations discard noise and complexity in order to give the human mind compact, usable descriptions. Second, scientific models undergo something like Darwinian selection—successful models survive and are published, while thousands of unsuccessful ideas disappear, making mathematics appear more consistently successful than it actually is.
He illustrates this with engineering, nonlinear systems, fractals, complex numbers and even the apparently elementary act of counting bananas. In the physical world, objects have fuzzy boundaries, measurements contain noise, and there are ultimately physical limits to how many objects can actually be counted or represented. Mathematical entities such as perfect circles, integers extending indefinitely, delta functions and exact real numbers therefore need not correspond to independently existing objects in nature. Abbott adopts a deliberately strong non-Platonist position: mathematical models should be regarded primarily as useful constructions rather than literal descriptions of what reality “really is.” This view has practical consequences, he argues, because treating mathematics as something humans are free to redesign may encourage better formalisms—for example, his advocacy of geometric algebra as a more unified alternative to the traditional mixture of scalars, vectors, complex numbers, quaternions, dot products and cross products.
Key takeaways
ARTICLE [PDF]
Author: Derek Abbott
Published: October 2013
Journal:Proceedings of the IEEE, Vol. 101, No. 10, pp. 2147–2153
DOI: 10.1109/JPROC.2013.2274907
Derek Abbott challenges Eugene Wigner’s famous claim about the “unreasonable effectiveness of mathematics” in describing nature. Rather than regarding the success of mathematics in physics as something mysterious or as evidence that mathematical objects have an independent Platonic existence, Abbott argues that mathematics is largely a human invention constructed to represent patterns and regularities that humans happen to observe. We naturally notice problems that our mathematical tools can solve, develop new mathematics when existing tools fail, and tend to remember successful mathematical models while overlooking the enormous number of unsuccessful ones. Thus, what appears to be a miraculous correspondence between mathematics and nature may partly be a form of selection bias.
Abbott develops Richard Hamming's earlier observations: we see what we are equipped to look for; we choose the mathematics appropriate to particular problems; science successfully addresses only a comparatively small subset of conceivable questions; and human cognition itself was shaped by evolution to understand phenomena occurring on approximately human spatial and temporal scales. Abbott adds two ideas of his own. First, mathematical laws function as a kind of lossy compression of reality: equations discard noise and complexity in order to give the human mind compact, usable descriptions. Second, scientific models undergo something like Darwinian selection—successful models survive and are published, while thousands of unsuccessful ideas disappear, making mathematics appear more consistently successful than it actually is.
He illustrates this with engineering, nonlinear systems, fractals, complex numbers and even the apparently elementary act of counting bananas. In the physical world, objects have fuzzy boundaries, measurements contain noise, and there are ultimately physical limits to how many objects can actually be counted or represented. Mathematical entities such as perfect circles, integers extending indefinitely, delta functions and exact real numbers therefore need not correspond to independently existing objects in nature. Abbott adopts a deliberately strong non-Platonist position: mathematical models should be regarded primarily as useful constructions rather than literal descriptions of what reality “really is.” This view has practical consequences, he argues, because treating mathematics as something humans are free to redesign may encourage better formalisms—for example, his advocacy of geometric algebra as a more unified alternative to the traditional mixture of scalars, vectors, complex numbers, quaternions, dot products and cross products.
Key takeaways
- Mathematics works, but not miraculously: its successes may look extraordinary because we disproportionately notice successful applications.
- Mathematical models are approximations: Abbott describes them as compressed, idealized representations of a noisy and complicated reality.
- Mathematics may be anthropocentric: our mathematics reflects human perception, evolution, cognitive limitations and the scales at which we live.
- The article argues against mathematical Platonism: Abbott concludes that mathematics is best viewed as a powerful human-created tool for describing regularities, not necessarily as a pre-existing structure of the universe.
ARTICLE [PDF]
┌────────────────────────────────┐
│ KONSTANTINOS MICHAILIDIS │
└────────────────────────────────┘
│ KONSTANTINOS MICHAILIDIS │
└────────────────────────────────┘