String-Gas Cosmology
The early universe was a hot, dense gas of tightly wound strings rather than a single point. String thermodynamics naturally produces our observed cosmology.
Placeholder for a 3D visualisation of Cyclic & Bouncing Cosmologies. The interactive scene will land in Phase 3. There is no absolute beginning. Our universe is the latest phase in an eternal cycle. Variants disagree on the mechanism: a conformal transition (Penrose), a quantum bounce (LQC), a brane collision (Ekpyrotic), or a dilaton-driven bounce (Pre-Big Bang).
In one sentence
String-Gas Cosmology proposes that the Big Bang was the dynamics of a gas of fundamental strings, with cosmic structure emerging from string thermodynamics.
The claim
In String-Gas Cosmology, the very early universe consisted of a hot gas of fundamental strings (the objects predicted by string theory). These strings were not infinitely long; they were wound, knotted, and interacting. The cosmological dynamics of this string gas is qualitatively different from the standard inflationary cosmology, leading to different predictions for cosmic structure.
Brandenberger and Vafa worked out the basic mathematical framework in 1989. The key insight: in string theory, the lowest-energy states are not point particles but wound strings. As the universe expands, certain modes of these strings 'unwind' and disappear, leaving the spatial dimensions we observe today. The model naturally explains why we see exactly three spatial dimensions: only three dimensions allowed strings to unwind sufficiently as the universe expanded.
More recent work has shown that String-Gas Cosmology can produce a nearly scale-invariant spectrum of cosmological perturbations, similar to what inflation predicts but through a different mechanism. The model is one of the more developed non-inflationary string theory cosmologies.
The family stance
A previous cycle, aeon, contracting phase, or alternate-brane state existed before our universe. The "before" is a physically connected predecessor, not nothing or another arena.
Predictions
- The number of large spatial dimensions is constrained by string dynamics (predicts three)
- Specific spectrum of cosmological perturbations from string thermodynamics
- Distinct signatures in CMB power spectrum that differ from inflation
- Relationship between primordial gravitational waves and density perturbations
Evidence
- Naturally explains why three spatial dimensions are large rather than all six (in the original 10D string theory framework)
- Provides a non-inflationary mechanism for the observed scale-invariant spectrum
- Connects directly to fundamental string theory rather than effective field theory
Counterpoints
- String theory itself has not been experimentally verified
- Specific predictions are model-dependent and not yet decisive
- The mechanism for emerging from the high-temperature string gas phase is not fully worked out
Variants in this family
▸Go deeperTechnical detail with proper terminology
The Brandenberger-Vafa 1989 paper formulates the model in 10-dimensional string theory. The crucial concept is the Hagedorn temperature, a maximum temperature above which the string gas description breaks down. Just below the Hagedorn temperature, strings dominate and the dynamics are exotic.
In subsequent work, Brandenberger and collaborators showed that as the universe expands and cools, three spatial dimensions naturally decompactify (grow large) while the other six remain compactified at the string scale. This decompactification is driven by the dynamics of winding modes that can annihilate only in three dimensions.
The model has been extended to make connections with observational cosmology. The Brandenberger group at McGill and others have computed predictions for the CMB power spectrum and gravitational wave background that can be compared to observations.
References
- EstablishedBrandenberger, R., Vafa, C. (1989). 'Superstrings in the early universe.' Nucl. Phys. B 316, 391
- EstablishedBrandenberger, R. (2006). 'String gas cosmology.' arXiv:hep-th/0608121
Last reviewed May 14, 2026
Spotted an error? Have a source to add?
Prefer email?
You can also send a prefilled email with the variant URL already filled in.