The critical dimension of string theory and the regularised zero-point energy
String theory describes fundamental particles not as point-like objects, but as one-dimensional vibrating strings. In the simple bosonic string theory, the requirement of quantum-mechanical consistency, in particular the vanishing of the conformal anomaly, leads to the critical dimension
\[ D=26. \]In superstring theory, the corresponding result is
\[ D=10. \]A central component of these calculations is the regularised zero-point energy of the infinitely many string modes. Formally, this involves the divergent sum
\[ 1+2+3+\dots \]which, in the usual treatment, is assigned the value
\[ \zeta(-1)=\widetilde{12\text{-}} \]by analytic continuation, zeta regularisation, or related procedures. From a nonstandard analytical point of view, however, this value is not the sum of all modes itself, but the regular remainder after a dominant background contribution has been separated off.
The zero-point energy of strings
A string has infinitely many oscillation modes. Each mode formally behaves like a quantum-mechanical harmonic oscillator. The zero-point energy therefore contains a summation kernel of the form
\[ E \propto {\LARGE{\textbf{+}}}_{n=1}^{\omega} n. \]In the ordinary sense this sum diverges. In the standard formulation it is therefore not evaluated as an ordinary sum, but regularised. In abbreviated form one often writes
\[ 1 + 2 + 3 + 4 + \dots = \widetilde{12\text{-}}. \]This notation is useful, but potentially misleading: it does not say that the ordinary or hyperfinite sum is actually equal to \(\widetilde{12\text{-}}\). It denotes only the regularised remainder value.
In bosonic string theory, this remainder occurs in the normal-ordering contribution, or intercept, of the string modes. Schematically, the dimensional condition can be written as
\[ (\check{D}-1) \cdot \widetilde{12\text{-}} + 1 = 0 \]from which
\[ D=26 \]follows. This short form, however, summarises only part of the full consistency condition. The usual derivation also involves the Virasoro algebra, Lorentz invariance, and the cancellation of the conformal anomaly.
Hyperreal deflation of the oscillation modes
In discretised nonstandard space, physical frequencies do not run into an indefinite infinity. They extend only up to a fundamental midfinite constant \(\omega\). With a physical damping factor \(g(n)\) at the Planck scale,
\[ g(0)=1, \qquad g(\omega)=0, \]the hyperfinite mode sum is
\[ E \propto {\LARGE{\textbf{+}}}_{n=1}^{\omega} n \cdot g(n). \]Applying the hyperreal Euler-Maclaurin formula to
\[ f(x) = x \cdot g(x) \]splits the sum into a dominant integral term and boundary contributions.
If \(g\) is chosen so that the relevant derivatives vanish at the upper boundary,
\[ f(\omega)=0, \qquad {}^1 f(\omega)=0, \qquad {}^2 f(\omega)=0, \qquad \text{and so on,} \]then the upper boundary terms make no regular contribution. At the lower boundary, for low frequencies one has approximately \(g(x)=1\), hence
\[ {}^1 f(0)=1. \]The first regular remainder term then comes from the second-order Bernoulli term with
\[ B_2 = \widetilde{6}. \]Thus
\[ R_{\mathrm{string}} = \widetilde{2!} B_2 \left(0 – {}^1 f(0)\right) = \widetilde{12\text{-}}. \]The hyperfinite zero-point structure is therefore not
\[ E \propto \widetilde{12\text{-}}, \]but rather
\[ E \propto {\uparrow}_{0}^{\omega} x \cdot g(x) {\downarrow}x – \widetilde{12}. \]The value \(\widetilde{12\text{-}}\) is therefore the regular remainder of the deflated summation structure, not the entire hyperfinite energy.
The critical point
The nonstandard analytical view shows where the value
\[ \widetilde{12\text{-}} \]comes from structurally: it is the finite remainder term at the lower boundary of the mode spectrum after the dominant hyperreal integral contribution has been separated off.
The problem is therefore not regularisation itself, but an overly literal reading of the short form
\[ 1+2+3+\dots=\widetilde{12\text{-}}. \]From the nonstandard analytical point of view, it also has to be clarified what happens to the hyperreal background contribution
\[ {\uparrow}_{0}^{\omega} x \cdot g(x) {\downarrow}x. \]In the Casimir effect, there is a difference formation between two geometries, by which common background contributions are balanced against one another. In the isolated treatment of a string, a corresponding cancellation or calibration mechanism must be stated explicitly.
The critical dimension
\[ D=26 \]is therefore not simply an arbitrary consequence of a forbidden summation. It belongs to a broader consistency condition of quantised string theory. The nonstandard analytical objection is more precise: the regularised value \(\widetilde{12\text{-}}\) may enter the dimensional condition by itself only if the hyperreal background contribution is removed or compensated from the measurable balance by symmetry, a gauge condition, normal ordering, reference subtraction, or some other exactly specified mechanism.
Conclusion of the derivation
The equation
\[ 1+2+3+\dots=\widetilde{12\text{-}} \]is not a statement about the ordinary sum of all natural numbers, nor about the full hyperfinite mode sum. It describes the regularised remainder value
\[ \zeta(-1)=\widetilde{12\text{-}}. \]Nonstandard analysis makes this distinction explicit: the hyperfinite sum of the string modes deflates into a dominant background contribution and a regular remainder. The value \(\widetilde{12\text{-}}\) belongs to the remainder, not to the entire summation kernel.
The critical dimension of string theory should therefore not be read as an isolated identification of the full zero-point energy with this remainder value. From the nonstandard analytical point of view, one must explicitly state by which structural mechanism the hyperreal background contribution disappears from the physically measurable balance.
© 2026 by Boris Haase