🔢 mathematics

The Parafree Conjecture for associative algebras

This paper disproves the analogue of the Parafree Conjecture for associative algebras by constructing a finitely generated parafree augmented associative algebra with countably infinite-dimensional second homology, thereby answering a question posed by Ivanov and Lopatkin.

Original authors: Vasily Ionin, Roman Mikhailov

Published 2026-08-24

📖 7 min read🧠 Deep dive

CC0 1.0

Original authors: Vasily Ionin, Roman Mikhailov

Original paper dedicated to the public domain under CC0 1.0 (http://creativecommons.org/publicdomain/zero/1.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

In the vast landscape of mathematics, there is a branch dedicated to the study of structures that follow specific rules of combination, much like how words combine to form sentences or numbers combine to form equations. Within this field, researchers often look for the simplest possible versions of these structures, known as free objects. These are the building blocks that have no hidden restrictions or extra rules binding them together; they are pure and unrestricted. For decades, mathematicians have been fascinated by a class of structures that look exactly like these free building blocks when examined through a specific, limited lens, yet might be different underneath. These are called parafree objects. They behave identically to free objects in every finite step of a certain process, making them nearly impossible to distinguish from the real thing without looking at the whole infinite picture. The big question has been whether these look-alikes are actually just free objects in disguise, or if they possess hidden complexities that only reveal themselves when one looks at the entire infinite structure. This inquiry is not just a game of abstract logic; it touches on the fundamental nature of symmetry and shape in mathematics, helping to define the boundaries between what is simple and what is complex.

A team of researchers has now constructed a specific example that settles a long-standing debate about these parafree structures within the realm of associative algebras, which are systems where you can multiply elements together in a fixed order. For years, a prevailing idea suggested that if such a structure is generated by a finite number of elements, it must be simple in a very specific way: its second layer of complexity, a measure of how the pieces fit together, should be empty. This idea, known as the Parafree Conjecture, implied that these look-alike structures could not have hidden, infinite depths of complexity if they started with a finite set of rules. The researchers set out to test this by building a new algebraic object from scratch. They began with a set of six basic elements and imposed a series of rules that linked them together in a pattern that stretched on forever. The rules were designed so that the first few layers of the structure would match a free system perfectly, but the infinite tail of the rules would introduce a subtle, persistent irregularity.

The result of their construction is a structure that is generated by a finite number of elements, yet it is not finitely presented, meaning it cannot be fully described by a finite list of rules. More importantly, the researchers proved that this object is parafree, behaving exactly like a free system in every finite approximation. However, when they examined the second layer of its complexity, they found it was not empty as the conjecture predicted. Instead, it was infinitely large, containing a countable infinity of independent pieces of information. This discovery definitively disproves the analogue of the Parafree Conjecture for associative algebras. It shows that a structure can be built from a finite number of starting points and mimic a free system perfectly in every finite test, yet still harbor an infinite reservoir of hidden complexity that only appears when the entire infinite structure is considered.

To understand how this works, imagine the structure as a tower built from blocks. The researchers started with a few types of blocks and a set of instructions for stacking them. The instructions were cleverly written so that if you only looked at the bottom ten layers, the tower looked exactly like a standard, free tower with no restrictions. But the instructions included a rule that applied to the hundredth layer, the thousandth, and every layer after that, creating a subtle mismatch that never resolved. This mismatch meant that while the tower looked free from the ground up to any specific height, the full tower contained an endless number of unique, non-repeating patterns that could not be simplified away. The researchers showed that this infinite complexity is real and measurable, existing in a specific mathematical space that counts how the pieces of the structure interlock.

The significance of this finding lies in what it reveals about the limits of finite descriptions. It demonstrates that knowing how a system behaves in every finite step is not enough to guarantee that the whole system is simple. The researchers used a method involving a sub-monoid of a free monoid, which is essentially a collection of words formed from a specific alphabet that follows certain concatenation rules. They identified a specific set of words that could be generated by a finite list of starting words but required an infinite list of rules to fully define. By translating these word rules into algebraic equations, they created the counterexample. The work confirms that the property of being "free" is not something that can be fully captured by looking at finite snapshots, even if those snapshots are perfect.

This result answers a question that had been posed by other mathematicians regarding the homological properties of these algebras. Homology, in this context, is a way of counting the holes or independent cycles within a structure. The researchers found that their constructed algebra has a second homology group that is infinite-dimensional. This means there are infinitely many independent ways the structure can loop back on itself that cannot be shrunk to nothing. This stands in stark contrast to the behavior of truly free algebras, which have no such loops. The construction proves that the class of parafree algebras is much richer and more complex than previously thought, containing objects that are finitely generated but infinitely complex in their internal connections.

The paper also provides a detailed map of how this complexity grows. It shows that the process of peeling away layers of the structure to reveal its core takes an infinite number of steps, specifically reaching a point that mathematicians describe as a transfinite length. This indicates that the structure is not just complex, but complex in a way that defies standard finite counting. The researchers did not just find a single example; they provided a blueprint for how such examples are constructed, showing that they arise naturally from the study of sub-monoids of free monoids. This connects the abstract world of algebra to the more concrete world of word combinations, showing that the rules governing how words can be formed can lead to deep algebraic surprises.

Ultimately, this work serves as a reminder that in mathematics, intuition based on finite cases can sometimes lead us astray when dealing with the infinite. The researchers have shown that a structure can wear the mask of simplicity perfectly, fooling every finite test, while hiding a vast, infinite interior. Their example is a concrete proof that the Parafree Conjecture, in the form it was proposed for associative algebras, is false. The field of algebraic structures now has a new, well-defined example of a finitely generated object that is parafree but not free, with a second homology group that is countably infinite. This finding closes one chapter of inquiry while opening new questions about the full range of behaviors possible in these systems, ensuring that the study of parafree objects will continue to be a vibrant area of mathematical research.

Technical Summary: The Parafree Conjecture for Associative Algebras

Problem Statement
The paper addresses the validity of the Parafree Conjecture within the category of augmented associative algebras. In group theory, a parafree group is defined as a residually nilpotent group that shares the same lower central series quotients as a free group. A central open question, known as the Parafree Conjecture, posits that for every finitely generated parafree group $G$ , the second homology group $H_2(G; \mathbb{Z})$ vanishes.

This conjecture has been extended to other algebraic categories, including Lie algebras and augmented associative algebras. Ivanov and Lopatkin [IL21] previously constructed a finitely generated parafree augmented associative algebra of infinite cohomological dimension, disproving the "strong" form of the conjecture (which asserts cohomological dimension $\le 2$ ). However, the specific question of whether $H_2(B) = 0$ for every finitely generated parafree augmented associative algebra $B$ remained open. This paper seeks to resolve that specific question.

Methodology
The authors construct a counterexample using the monoid algebra of a specific submonoid of a free monoid. The methodology proceeds through the following steps:

Construction of the Algebra: The authors define an augmented algebra $A$ over a field $k$ generated by six elements $\{a, b, c, d, x, y\}$ subject to an infinite set of relations derived from the identity:
$e_\ell = (1 + a)(1 + b)^\ell(1 + c) - (1 + d)(1 + x)^\ell(1 + y) = 0, \quad \ell \ge 0.$
This algebra is identified as the monoid algebra $k[L]$ of a submonoid $L$ of a free monoid, a structure known to be finitely generated but not finitely presented (based on Spehner's construction [Sp77]).
Auxiliary Algebra and Para-equivalence: To analyze the homological properties, the authors introduce an auxiliary algebra $A_0$ defined by the same generators but only the first two relations ( $e_0, e_1$ ). They prove that the canonical map from the free algebra on four generators, $k\langle a, b, c, d \rangle$ , to $A_0$ is a para-equivalence (an isomorphism on all nilpotent quotients). This establishes that $A_0$ is parafree of rank four.
Homological Computation via Gröbner–Shirshov Bases: The authors employ Anick's resolution and Gröbner–Shirshov basis theory to compute the second homology group $H_2(A)$ . By establishing that the family of relations $\{f_\ell\}$ (linearly equivalent to $\{e_\ell\}$ ) forms a reduced Gröbner–Shirshov basis with specific maximal terms, they demonstrate that the kernel of the map $A_0 \to A$ corresponds to the intersection of the powers of the augmentation ideal of $A_0$ , denoted $I(A_0)^\omega$ .
Transfinite Augmentation Series: The paper analyzes the transfinite powers of the augmentation ideal $I(A_0)$ . Using ordinal-indexed powers, the authors show that the series $\{I(A_0)_\alpha\}$ has a transfinite length of $\omega^2$ , implying that the intersection of finite powers $I(A_0)^\omega$ is non-zero (specifically, it equals the kernel of the projection $A_0 \to A$ ), while the intersection of the transfinite series eventually vanishes at $\omega^2$ .

Key Results
The paper establishes the following main theorems:

Theorem A: The augmented algebra $A$ is parafree of rank four. However, its second homology group is countably infinite-dimensional:
$H_2(A) \cong k^{\oplus \aleph_0}.$
Furthermore, $A$ is finitely generated but not finitely presented.
Theorem B: The canonical map $k\langle a, b, c, d \rangle \to A_0$ is a para-equivalence. The kernel of the projection $A_0 \to A$ is exactly the intersection of the powers of the augmentation ideal of $A_0$ , denoted $I(A_0)^\omega$ . The second homology $H_2(A)$ is isomorphic to the quotient:
$\frac{I(A_0)^\omega}{I(A_0) \cdot I(A_0)^\omega + I(A_0)^\omega \cdot I(A_0)} \cong \bigoplus_{\ell \ge 2} k.$
A basis for this homology is represented by the classes of the elements $(1+a)b^\ell(1+c) - (1+d)x^\ell(1+y)$ for $\ell \ge 2$ .
Theorem C: The transfinite augmentation series $\{I(A_0)_\alpha\}$ has a transfinite length of $\omega^2$ . Specifically, $I(A_0)_{\omega^2} = 0$ , but for any ordinal $\alpha < \omega^2$ , $I(A_0)_\alpha \neq I(A_0)_{\alpha+1}$ .

Significance and Conclusion
The primary contribution of this work is the disproof of the associative-algebra analogue of the Parafree Conjecture. The authors demonstrate that the property $H_2(B) = 0$ does not hold for all finitely generated parafree augmented associative algebras, even when the algebra is finitely generated.

The paper clarifies that while parafree algebras share the nilpotent quotients of free algebras, they do not necessarily share the homological finiteness properties (specifically the vanishing of $H_2$ ) of free algebras. The counterexample provided is the monoid algebra of a finitely generated, non-finitely presented submonoid of a free monoid. This result answers a specific question posed by Ivanov and Lopatkin regarding the vanishing of the second homology group in this category. The work relies on the interplay between monoid theory, Gröbner–Shirshov bases, and the homological algebra of augmented algebras, without proposing further applications or extensions beyond the resolution of the stated conjecture.

Technical Summary: The Parafree Conjecture for Associative Algebras

More like this