Monads and Algebras over a Monad

⚡ Key Takeaways

A monad on a category $\mathcal{C}$ is a monoid in the category of endofunctors: an endofunctor $T$ equipped with a unit $\eta\colon \mathrm{Id}\Rightarrow T$ and multiplication $\mu\colon T^2\Rightarrow T$ satisfying associativity and unit laws. An algebra over the monad (Eilenberg–Moore algebra) is an object $A$ with a structure map $a\colon TA\to A$ compatible with $\eta$ and $\mu$ . Every adjunction generates a monad, and every monad arises from an adjunction in two canonical ways — via the Kleisli and Eilenberg–Moore categories. Applied across universal algebra, topology, functional programming, and descent theory. Originated by Godement (1958), Huber, Kleisli, Eilenberg–Moore, and Beck (1960s).

💡 The Core Insight

Adjunctions and monads are two faces of the same phenomenon. An adjunction $F\dashv U$ between $\mathcal{C}$ and $\mathcal{D}$ produces a monad $T = UF$ on $\mathcal{C}$ that remembers the composite round-trip and forgets the intermediate category $\mathcal{D}$ . The monad alone cannot reconstruct $\mathcal{D}$ , but it can reconstruct two canonical "best approximations": the Kleisli category $\mathcal{C}_T$ (the initial adjunction giving $T$ ) and the Eilenberg–Moore category $\mathcal{C}^T$ (the terminal one). All adjunctions inducing $T$ sit between these extremes, and the original $\mathcal{D}$ is monadic exactly when the comparison functor $\mathcal{D}\to\mathcal{C}^T$ is an equivalence.

DefinitionMonad

A monad on a category $\mathcal{C}$ is a triple $(T,\eta,\mu)$ where $T\colon\mathcal{C}\to\mathcal{C}$ is an endofunctor, $\eta\colon\mathrm{Id}_{\mathcal C}\Rightarrow T$ (the unit), and $\mu\colon T\circ T\Rightarrow T$ (the multiplication) are natural transformations satisfying the associativity and left/right unit coherence laws.

1

\quad \mu \circ T\mu \;=\; \mu \circ \mu T \quad\text{as natural transformations } T^3\Rightarrow T

Derivation hint: Both sides are the two ways to collapse $TTT \Rightarrow T$ ; equality is the monoid associativity condition transported to $\mathrm{End}(\mathcal C)$ under horizontal composition $\circ$ .

2

\quad \mu \circ \eta T \;=\; \mathrm{id}_T \;=\; \mu \circ T\eta

Derivation hint: Insert a unit on the left or right factor of $TT$ and multiply; this is the monoid unit law $e\cdot x = x = x\cdot e$ for the monoid $(T,\mu,\eta)$ . Here $\eta T, T\eta\colon T\Rightarrow T^2$ are the whiskerings: $(\eta T)_X=\eta_{TX}$ and $(T\eta)_X = T(\eta_X)$ . Likewise $(\mu T)_X=\mu_{TX}$ and $(T\mu)_X = T(\mu_X)$ . The diagrams asserted componentwise are, for every object $X$ :

3

\quad \mu_X \circ T(\mu_X) = \mu_X \circ \mu_{TX} \colon T^3X\to TX, \qquad \mu_X\circ\eta_{TX} = \mathrm{id}_{TX} = \mu_X\circ T(\eta_X)

Derivation hint: Evaluate $[1]$ and $[2]$ at an object $X$ ; the symbols $T\mu, \mu T, \eta T, T\eta$ unwind to the indicated component morphisms.

DefinitionEilenberg–Moore Algebra over a Monad

Given a monad $(T,\eta,\mu)$ , a $T$ -algebra (or Eilenberg–Moore algebra) is a pair $(A,a)$ with $A\in\mathcal C$ and a morphism $a\colon TA\to A$ (the structure map) satisfying:

4

\quad a\circ \eta_A = \mathrm{id}_A \qquad(\text{unit law})

Derivation hint: A "generator" $x\in A$ , embedded via $\eta_A$ into the free structure $TA$ , must evaluate back to itself.

5

\quad a\circ \mu_A = a\circ T(a) \qquad(\text{associativity / coherence law})

Derivation hint: Evaluating a doubly-nested formal expression $T(TA)$ either by first multiplying ( $\mu_A$ ) then evaluating, or by first evaluating the inner layer ( $Ta$ ) then the outer, gives the same result — exactly the associativity of an action. A morphism of $T$ -algebras $f\colon (A,a)\to(B,b)$ is a $\mathcal C$ -morphism $f\colon A\to B$ with $f\circ a = b\circ Tf$ . These form the Eilenberg–Moore category $\mathcal C^T$ , with forgetful functor $U^T\colon\mathcal C^T\to\mathcal C$ , $(A,a)\mapsto A$ .

DefinitionKleisli Category

The Kleisli category $\mathcal C_T$ has the same objects as $\mathcal C$ . A morphism $X\rightsquigarrow Y$ in $\mathcal C_T$ is a $\mathcal C$ -morphism $f\colon X\to TY$ . The identity on $X$ is $\eta_X$ , and Kleisli composition of $f\colon X\to TY$ and $g\colon Y\to TZ$ is:

6

\quad g\diamond f \;=\; \mu_Z \circ T(g)\circ f \colon X\to TZ

Derivation hint: Lift $g$ to act on $TY$ via $T(g)\colon TY\to T^2Z$ , apply $f$ first, then flatten $T^2Z\to TZ$ with $\mu_Z$ ; associativity of $\diamond$ follows from $[1]$ , units from $[2]$ .

TheoremAdjunctions Generate Monads

Let $F\colon\mathcal C\to\mathcal D$ and $U\colon\mathcal D\to\mathcal C$ with $F\dashv U$ , unit $\eta\colon\mathrm{Id}\Rightarrow UF$ and counit $\varepsilon\colon FU\Rightarrow\mathrm{Id}$ . Then $T=UF$ with the given $\eta$ and with $\mu = U\varepsilon F\colon UFUF\Rightarrow UF$ is a monad on $\mathcal C$ .

7

\quad T = UF,\qquad \eta = \eta,\qquad \mu = U\varepsilon F

Derivation hint: The triangle identities $\varepsilon F\circ F\eta=\mathrm{id}_F$ and $U\varepsilon\circ\eta U=\mathrm{id}_U$ , whiskered appropriately, yield the monad unit laws $[2]$ ; the naturality of $\varepsilon$ yields associativity $[1]$ .

TheoremEilenberg–Moore and Kleisli Adjunctions; Comparison

Every monad $T$ on $\mathcal C$ arises from at least two adjunctions:

The Eilenberg–Moore adjunction $F^T\dashv U^T$ where $F^T X=(TX,\mu_X)$ is the free $T$ -algebra and $U^T(A,a)=A$ , with $U^TF^T=T$ .
The Kleisli adjunction $F_T\dashv U_T$ between $\mathcal C$ and $\mathcal C_T$ , with $U_TF_T=T$ .

Among all adjunctions inducing $T$ , the Kleisli adjunction is initial and the Eilenberg–Moore adjunction is terminal in the category $\mathrm{Adj}(T)$ of such adjunctions. There is a fully faithful comparison functor $K\colon\mathcal C_T\to\mathcal C^T$ exhibiting $\mathcal C_T$ as the full subcategory of free algebras, and for any adjunction $F\dashv U$ inducing $T$ a unique comparison functor $\Phi\colon\mathcal D\to\mathcal C^T$ with $U^T\Phi=U$ and $\Phi F=F^T$ .

8

\quad \Phi(D) = \big(U D,\; U\varepsilon_D\colon TUD = UFUD\to UD\big)

Derivation hint: $UD$ carries a $T$ -action because the counit at $D$ gives $\varepsilon_D\colon FUD\to D$ ; applying $U$ produces a structure map; the algebra laws $[4],[5]$ are the triangle identity and the naturality of $\varepsilon$ .

ProofAdjunction ⇒ Monad (verification of the laws)

Let $F\dashv U$ with unit $\eta$ , counit $\varepsilon$ , satisfying the triangle identities

9

\quad \varepsilon F\circ F\eta=\mathrm{id}_F,\qquad U\varepsilon\circ\eta U=\mathrm{id}_U.

Derivation hint: These are the defining zig-zag equations of any adjunction, expressed with whiskering.

Step 1 — Set $T=UF$ , $\mu=U\varepsilon F$ . Then $\mu$ is a natural transformation $T^2=UFUF\Rightarrow UF=T$ since $\varepsilon\colon FU\Rightarrow\mathrm{Id}_{\mathcal D}$ is natural and $U,F$ are functors; whiskering preserves naturality.

Step 2 — Right unit $\mu\circ T\eta=\mathrm{id}_T$ . Compute $T\eta = UF\eta\colon UF\Rightarrow UFUF$ . Then

\mu\circ T\eta = U\varepsilon F\circ UF\eta = U(\varepsilon F\circ F\eta) = U(\mathrm{id}_F)=\mathrm{id}_{UF}=\mathrm{id}_T,

using functoriality of $U$ and the first triangle identity of $[9]$ .

Step 3 — Left unit $\mu\circ\eta T=\mathrm{id}_T$ . Here $\eta T=\eta UF\colon UF\Rightarrow UFUF$ . Then

\mu\circ\eta T = U\varepsilon F\circ \eta UF = (U\varepsilon\circ\eta U)F = (\mathrm{id}_U)F=\mathrm{id}_{UF}=\mathrm{id}_T,

using the second triangle identity. (Whiskering by $F$ on the right is functorial.)

Step 4 — Associativity $\mu\circ T\mu=\mu\circ\mu T$ . Both sides are transformations $UFUFUF\Rightarrow UF$ . We have $T\mu=UF\,U\varepsilon F=U(FU\varepsilon)F$ and $\mu T=U\varepsilon F\,UF=U(\varepsilon FU)F$ . The interchange law applied to the naturality square of $\varepsilon$ at the morphism $\varepsilon$ itself — i.e. $\varepsilon\circ FU\varepsilon=\varepsilon\circ\varepsilon FU$ as transformations $FUFU\Rightarrow\mathrm{Id}$ — gives, after whiskering by $U$ on the left and $F$ on the right, exactly $\mu\circ T\mu=\mu\circ\mu T$ . $\qquad\blacksquare$

∎

ProofFree

T

-algebras and the unit/counit of

F^T\dashv U^T

Define $F^T X=(TX,\mu_X)$ . It is an algebra: the unit law $[4]$ for it reads $\mu_X\circ\eta_{TX}=\mathrm{id}_{TX}$ , which is the left half of $[3]$ ; the associativity law $[5]$ reads $\mu_X\circ\mu_{TX}=\mu_X\circ T\mu_X$ , which is the left half of $[1]$ evaluated at $X$ . So $(TX,\mu_X)$ is a genuine $T$ -algebra.

Adjunction data. The unit of $F^T\dashv U^T$ at $X$ is $\eta_X\colon X\to U^TF^TX=TX$ . The counit at an algebra $(A,a)$ is the algebra map

10

\quad \varepsilon^T_{(A,a)} = a\colon (TA,\mu_A)\to (A,a),

Derivation hint: $a$ is an algebra morphism because $[5]$ says $a\circ\mu_A=a\circ Ta$ , exactly the morphism condition $a\circ\mu_A = a\circ T(a)$ . The triangle identities for $F^T\dashv U^T$ are $a\circ\eta_A=\mathrm{id}_A$ (which is $[4]$ ) and $\mu_X\circ T\eta_X=\mathrm{id}_{TX}$ (right half of $[3]$ ). Hence $F^T\dashv U^T$ and $U^TF^T=T$ with the original $\eta,\mu$ . $\qquad\blacksquare$

∎

Figure

The two universal resolutions of a monad: Kleisli (initial) and Eilenberg–Moore (terminal)

mathematician · Jonathan Mock Beck | 1935–2006

Beck's 1967 Columbia thesis crystallized monad theory into a working tool. He proved the monadicity (tripleability) theorem, giving a sharp criterion for when a forgetful functor exhibits its source as the algebras of a monad, and he introduced distributive laws for composing monads. The associated Beck–Chevalley condition and Beck's bar/cobar resolution (the simplicial object built from $T$ ) underpin descent theory and monadic cohomology. The term "triple" — Beck's and Eilenberg's preferred word for "monad" — survives in the older literature; Mac Lane popularized "monad" by analogy with the monoid law.

🧠 Intuition

Think of $T$ as "the free gadget on a set of generators." For the monad on Set sending $X\mapsto$ (underlying set of the free monoid $X^*$ ), an element of $TX$ is a formal word in the generators $X$ . The unit $\eta_X\colon X\to TX$ injects a generator as a length-one word. The multiplication $\mu_X\colon T(TX)\to TX$ takes a "word of words" and concatenates — flattening one layer of nesting. The associativity law says flattening a triply-nested expression is unambiguous; the unit laws say a single generator wrapped trivially flattens back to itself. An algebra $a\colon TA\to A$ then says: " $A$ knows how to evaluate formal words into honest elements," coherently. For the monoid monad, an algebra is exactly a monoid: $a$ reads off the product of any formal word. The slogan is: a monad is a theory of formal expressions; an algebra is a model of that theory. The Kleisli category is the world of "programs that produce $T$ -effects," while the Eilenberg–Moore category is the world of "all models," and free models embed into it.

RG

Roger Godement

1921–2016

Though a number theorist and functional analyst, Godement introduced in his 1958 Topologie algébrique et théorie des faisceaux the standard construction (now the "monad of a comonad" / the Godement resolution): given a sheaf, the canonical flasque resolution iterates an endofunctor with exactly the comonad/monad coherences. This was the first systematic appearance of the data $(T,\eta,\mu)$ as a computational device, used to define sheaf cohomology via the Godement (bar) resolution. The pattern — build a simplicial object from iterated $T$ and take cohomology — became the template for monadic cohomology of all algebraic theories.

ExampleThe power-set monad and its algebras = complete sup-lattices

Given. On Set, let $T=\mathcal P$ (covariant power set), $\eta_X(x)=\{x\}$ , and $\mu_X\colon \mathcal P\mathcal P X\to \mathcal P X$ , $\mu_X(\mathcal S)=\bigcup_{S\in\mathcal S}S$ .

Step 1 — Check monad laws. Unit: $\mu_X(\eta_{\mathcal P X}(S))=\bigcup\{S\}=S$ and $\mu_X(\mathcal P\eta_X(S))=\bigcup_{x\in S}\{x\}=S$ , giving $[3]$ . Associativity: for $\mathcal T\in\mathcal P^3X$ ,

\mu_X(\mu_{\mathcal P X}(\mathcal T))=\bigcup_{\mathcal S\in\mathcal T}\bigcup_{S\in\mathcal S}S=\bigcup_{\mathcal S\in\mathcal T}\mu_X(\mathcal S)=\mu_X(\mathcal P\mu_X(\mathcal T)),

confirming $[1]$ .

Step 2 — Decode an algebra. A structure map $a\colon\mathcal P A\to A$ with $a(\{x\})=x$ (unit $[4]$ ) and $a(\bigcup_{S\in\mathcal S}S)=a(\{a(S):S\in\mathcal S\})$ (coherence $[5]$ ). Setting $\bigvee S := a(S)$ makes $a$ a complete join operation: $a(\emptyset)$ is a bottom element, $a$ is monotone and idempotent on the induced order.

Step 3 — Identify the category. Algebra morphisms are exactly sup-preserving maps.

\boxed{\ \mathbf{Set}^{\mathcal P}\;\simeq\;\mathbf{CSLat}_{\vee}\ \text{(complete join-semilattices with sup-preserving maps)}\ }

The free algebra on a set $X$ is $(\mathcal P X,\bigcup)$ , the power set ordered by inclusion.

ExampleThe free-vector-space monad and its algebras = vector spaces; a numeric Kleisli computation

Given. Fix a field $k$ . On Set, $T X = \{\text{finite formal }k\text{-combinations of elements of }X\}=k^{(X)}$ . Unit $\eta_X(x)=1\cdot x$ . Multiplication $\mu_X$ substitutes a combination-of-combinations and re-collects coefficients.

Step 1 — Algebra = $k$ -vector space. A structure map $a\colon k^{(A)}\to A$ assigns to a formal combination $\sum c_i a_i$ its actual value $\sum c_i a_i\in A$ . Laws $[4],[5]$ force $A$ to carry a $k$ -vector-space structure and morphisms to be $k$ -linear; thus $\mathbf{Set}^T\simeq \mathbf{Vect}_k$ .

Step 2 — A concrete Kleisli composite over $k=\mathbb Q$ . Take $X=\{x\},\;Y=\{y_1,y_2\},\;Z=\{z\}$ , and Kleisli arrows

f\colon X\to TY,\quad f(x)=2y_1+3y_2,\qquad g\colon Y\to TZ,\quad g(y_1)=5z,\; g(y_2)=-1z.

By $[6]$ , $g\diamond f = \mu_Z\circ T(g)\circ f$ . First $T(g)$ applied to $2y_1+3y_2$ gives the element-of- $TTZ$ " $2\,(5z)+3\,(-z)$ "; then $\mu_Z$ collects:

(g\diamond f)(x)=\mu_Z\big(2\cdot 5z+3\cdot(-1)z\big)=(10-3)\,z=7z.

Step 3 — Sanity check via matrices. Linear maps $\mathbb Q\!\xrightarrow{(2,3)^{\!\top}}\mathbb Q^2\xrightarrow{(5,\,-1)}\mathbb Q$ compose to the $1\times1$ matrix $5\cdot2+(-1)\cdot3=7$ .

\boxed{\ (g\diamond f)(x)=7z,\qquad \mathbf{Set}^T\simeq\mathbf{Vect}_{\mathbb Q}\ }

⚠️ Common mistake

A monad is not the same as its Kleisli category, nor as its algebras. The data $(T,\eta,\mu)$ on $\mathcal C$ determines both $\mathcal C_T$ and $\mathcal C^T$ , but these are generally very different categories: $\mathcal C_T$ is (equivalent to) the full subcategory of free algebras inside $\mathcal C^T$ . Confusing "model" (object of $\mathcal C^T$ ) with "computation" (morphism of $\mathcal C_T$ ) is the single most common error.

Also: the unit and multiplication are natural transformations, not arbitrary morphisms; their naturality squares are needed in nearly every proof. And the monad laws are about $T$ , $\eta$ , $\mu$ — not about any particular adjunction — so different adjunctions can yield isomorphic monads while having inequivalent middle categories $\mathcal D$ .

CounterexampleNot every endofunctor with "obvious" unit/mult is a monad; and not every adjunction-free

T

has algebras = something familiar

Consider $T X = X\times X$ on Set with $\eta_X(x)=(x,x)$ . There is no natural $\mu\colon (X\times X)\times(X\times X)\to X\times X$ making this a monad: any attempt (e.g. project to the diagonal pairs) violates the unit law $[2]$ , because $\mu\circ\eta T$ and $\mu\circ T\eta$ cannot both equal the identity for a multiplication that discards coordinates. Concretely, one checks for $|X|\ge 2$ that no natural transformation of the right type satisfies $[3]$ . A subtler point: the continuation monad $T X = R^{(R^X)}$ on Set is a monad and is monadic, but $\mathbf{Set}^T$ is not a category of finitary algebraic structures — it is not a variety in the sense of universal algebra, because $T$ does not preserve filtered colimits. This shows "Eilenberg–Moore category" is strictly more general than "category of algebras of an equational theory."

connection

Monads sit at a crossroads of category theory. They are precisely monoids in $\mathrm{End}(\mathcal C)$ , linking them to Monoidal Categories and the general theory of monoid objects. Dualizing every arrow gives Comonads and Coalgebras, whose coalgebras model coinductive/state-based systems. The relationship $F\dashv U\rightsquigarrow T$ is the categorical heart of Adjoint Functors, and the question of whether $\mathcal D\simeq\mathcal C^T$ is settled by Beck's Monadicity Theorem. In algebra, finitary monads on Set correspond exactly to Lawvere Theories and to varieties of universal algebra, while in semantics monads model Computational Effects (Moggi). Composing monads requires Distributive Laws of Monads, and the simplicial bar construction from $T$ feeds Monadic Cohomology and Descent.

🔬 Where this appears

⏳ History

- 1958 — Roger Godement introduces the standard construction (the comonad/monad of a resolution) in Topologie algébrique et théorie des faisceaux; first appearance of the

(T,\eta,\mu)

coherences in practice.

- 1958–1960 — Peter Huber, studying homotopy and adjunctions, observes that every adjunction yields the structure now called a monad; this circulates among Eilenberg's collaborators.

- 1962 — Samuel Eilenberg and J. C. Moore, in Adjoint functors and triples, define triples (= monads) abstractly and construct the Eilenberg–Moore category

\mathcal C^T

of algebras; prove the free–forgetful adjunction

F^T\dashv U^T

.

- 1965 — Heinrich Kleisli constructs the Kleisli category

\mathcal C_T

, exhibiting every monad as arising from an adjunction in a second, initial way.

- 1965 — Fred Linton connects finitary monads on Set with equational theories, foreshadowing the Lawvere-theory correspondence.

- 1967 — Jonathan Beck's Columbia thesis proves the monadicity (tripleability) theorem and introduces distributive laws and the bar resolution; the Beck–Chevalley condition emerges.

- 1969 — Saunders Mac Lane's Categories for the Working Mathematician adopts and popularizes the name monad and organizes the theory (Chapter VI).

- 1976 — Ernest Manes proves compact Hausdorff spaces = algebras of the ultrafilter monad, a striking non-equational monadicity result.

- 1991 — Eugenio Moggi's Notions of computation and monads imports monads into programming-language semantics; soon adopted by Wadler and the Haskell community, making monads a mainstream computational tool.

open-problem

Several structural questions remain technically delicate or actively studied. (1) Monadicity beyond the elementary toposes: characterizing exactly which functors between higher categories are monadic (the $\infty$ -categorical monadicity theorem of Lurie) and how monadic descent behaves in the $\infty$ -setting is still being worked out. (2) Composing monads: two monads $S,T$ compose to a monad $TS$ only given a distributive law $\lambda\colon ST\Rightarrow TS$ satisfying four coherence laws; deciding whether a distributive law exists — and classifying them, including "no-go" theorems (e.g. there is no distributive law combining the list monad with itself in certain ways) — is subtle and has yielded surprising negative results. (3) Algebraic vs. non-algebraic effects: which monads correspond to algebraic theories with operations and equations (Plotkin–Power) versus genuinely non-finitary effects (continuations) is a fine boundary; effect handlers exploit exactly the algebraic ones. (4) Codensity monads and ultrafilters: the codensity monad of the inclusion $\mathbf{FinSet}\hookrightarrow\mathbf{Set}$ is the ultrafilter monad (Kennison–Gildenhuys, Leinster), and the general theory of codensity monads as "limits over a for

Formal Statement

Why It Holds / Derivation

Building Intuition

Worked Examples

Common Misconceptions

Connections

Real-World Applications

Historical Development

Subtleties & Limits