
Let $\mathcal{P}({\bf N})$ be the power set of ${\bf N}$. We say that a
function $\mu^\ast: \mathcal{P}({\bf N}) \to \bf R$ is an upper density if, for
all $X,Y\subseteq{\bf N}$ and $h, k\in{\bf N}^+$, the following hold: (F1)
$\mu^\ast({\bf N}) = 1$; (F2) $\mu^\ast(X) \le \mu^\ast(Y)$ if $X \subseteq Y$;
(F3) $\mu^\ast(X \cup Y) \le \mu^\ast(X) + \mu^\ast(Y)$; (F4) $\mu^\ast(k\cdot
X) = \frac{1}{k} \mu^\ast(X)$, where $k \cdot X:=\{kx: x \in X\}$; (F5)
$\mu^\ast(X + h) = \mu^\ast(X)$.
We show that the upper asymptotic, upper logarithmic, upper Banach, upper
Buck, upper Polya, and upper analytic densities, together with all upper
$\alpha$densities (with $\alpha$ a real parameter $\ge 1$), are upper
densities in the sense of our definition. Moreover, we establish the mutual
independence of axioms (F1)(F5), and we investigate various properties of
upper densities (and related functions) under the assumption that (F2) is
replaced by the weaker condition that $\mu^\ast(X)\le 1$ for every
$X\subseteq{\bf N}$.
Overall, this allows us to extend and generalize results so far independently
derived for some of the classical upper densities mentioned above, thus
introducing a certain amount of unification into the theory.

Let $H$ be a multiplicatively written monoid. Given $k\in{\bf N}^+$, we
denote by $\mathscr U_k$ the set of all $\ell\in{\bf N}^+$ such that $a_1\cdots
a_k=b_1\cdots b_\ell$ for some atoms $a_1,\ldots,a_k,b_1,\ldots,b_\ell\in H$.
The sets $\mathscr U_k$ are one of the most fundamental invariants studied in
the theory of nonunique factorization, and understanding their structure is a
basic problem in the field: In particular, it is known that, in many cases of
interest, these sets are almost arithmetic progressions with the same
difference and bound for all large $k$, namely, $H$ satisfies the Structure
Theorem for Unions. The present paper improves the current state of the art on
this problem.
More precisely, we show that, under mild assumptions on $H$, not only does
the Structure Theorem for Unions hold, but there also exists $\mu\in{\bf N}^+$
such that, for every $M\in{\bf N}$, the sequences $$ \bigl((\mathscr
U_k\inf\mathscr U_k)\cap[\![0,M]\!]\bigr)_{k\ge 1} \quad\text{and}\quad
\bigl((\sup\mathscr U_k\mathscr U_k)\cap[\![0,M]\!]\bigr)_{k\ge 1} $$ are
$\mu$periodic from some point on. The result applies, e.g., to (the
multiplicative monoid of) all commutative Krull domains (e.g., Dedekind
domains) with finite class group; a variety of weakly Krull commutative domains
(including all orders in number fields with finite elasticity); some maximal
orders in central simple algebras over global fields; and all numerical
monoids.
Large parts of the proofs are worked out in a "purely additive model", by
inquiring into the properties of what we call a subadditive family, i.e., a
collection $\mathscr L$ of subsets of $\bf N$ such that, for all
$L_1,L_2\in\mathscr L$, there is $L\in\mathscr L$ with $L_1+L_2\subseteq L$.

We extend a few fundamental aspects of the classical theory of nonunique
factorization, as presented in Geroldinger and HalterKoch's 2006 monograph on
the subject, to a noncommutative and noncancellative setting, in the same
spirit of Baeth and Smertnig's work on the factorization theory of
noncommutative, but cancellative monoids [J. Algebra 441 (2015), 475551].
Then, we bring in power monoids and, applying the abstract machinery developed
in the first part, we undertake the study of their arithmetic.
More in particular, let $H$ be a multiplicatively written monoid. The set
$\mathcal P_{\rm fin}(H)$ of all nonempty finite subsets of $H$ is naturally
made into a monoid, which we call the power monoid of $H$ and is
noncancellative unless $H$ is trivial, by endowing it with the operation
$(X,Y) \mapsto \{xy: (x,y) \in X \times Y\}$. Power monoids are, in disguise,
one of the primary objects of interest in arithmetic combinatorics, and here
for the first time we tackle them from the perspective of factorization theory.
Proofs lead to consider various properties of finite subsets of $\mathbf N$
that can or cannot be split into a sumset in a nontrivial way, which gives
rise to a rich interplay with additive number theory.

Let $H$ be a multiplicatively written monoid (in particular, a group). We
denote by $\mathcal P_{\rm fin,\times}(H)$ the monoid obtained by endowing the
collection of all finite subsets of $H$ containing a unit of $H$ with the
operation of setwise multiplication $(X,Y) \mapsto \{xy: x \in X, y \in Y\}$,
and we study fundamental features of the arithmetic of this structure.
In particular, we show that $\mathcal{P}_{\text{fin},\times}(H)$ is atomic
(that is, each nonunit is a product of irreducibles) if and only if $H$ has no
nonidentity element $x$ with $x^2=1$ or $x^2 = x$. As a corollary, we obtain
that $\mathcal{P}_{\text{fin},\times}(H)$ is BF if and only if $H$ is
torsionfree. Also, we introduce a suitable notion of "minimal factorization"
to account for the fact that $\mathcal P_{\text{fin},\times}(H)$ may contain
nontrivial idempotents, in which case standard definitions from Factorization
Theory degenerate and lose their significance. Accordingly, we characterize
when $\mathcal{P}_{\text{fin},\times}(H)$ is a "minimal BFmonoid"
(respectively, "minimal HFmonoid"), i.e., each of its elements has minimal
factorizations of bounded (respectively, unique) length.
Many proofs boil down to considering sumset decompositions in
$(\mathbf{Z}/n\mathbf{Z}, +)$ or $(\mathbf N, +)$, which gives rise to an
intriguing interplay with Additive Combinatorics.

Given $k, \ell \in {\bf N}^+$, let $x_{i,j}$ be, for $1 \le i \le k$ and $0
\le j \le \ell$, some fixed integers, and define, for every $n \in {\bf N}^+$,
$s_n := \sum_{i=1}^k \prod_{j=0}^\ell x_{i,j}^{n^j}$. We prove that the
following are equivalent:
(a) There are a real $\theta > 1$ and infinitely many indices $n$ for which
the number of distinct prime factors of $s_n$ is greater than the
superlogarithm of $n$ to base $\theta$.
(b) There do not exist nonzero integers $a_0,b_0,\ldots,a_\ell,b_\ell $ such
that $s_{2n}=\prod_{i=0}^\ell a_i^{(2n)^i}$ and $s_{2n1}=\prod_{i=0}^\ell
b_i^{(2n1)^i}$ for all $n$.
We will give two different proofs of this result, one based on a theorem of
Evertse (yielding, for a fixed finite set of primes $\mathcal S$, an effective
bound on the number of nondegenerate solutions of an $\mathcal S$unit
equation in $k$ variables over the rationals) and the other using only
elementary methods.
As a corollary, we find that, for fixed $c_1, x_1, \ldots,c_k, x_k \in
\mathbf N^+$, the number of distinct prime factors of $c_1 x_1^n+\cdots+c_k
x_k^n$ is bounded, as $n$ ranges over $\mathbf N^+$, if and only if
$x_1=\cdots=x_k$.

Let $(X, \mathscr{L}, \lambda)$ and $(Y, \mathscr{M}, \mu)$ be finite measure
spaces for which there exist $A \in \mathscr{L}$ and $B \in \mathscr{M}$ with
$0 < \lambda(A) < \lambda(X)$ and $0 < \mu(B) < \mu(Y)$, and let $I\subseteq
\mathbf{R}$ be a nonempty interval. We prove that, if $f$ and $g$ are
continuous bijections $I \to \mathbf{R}^+$, then the equation $$
f^{1}\!\left(\int_X f\!\left(g^{1}\!\left(\int_Y g \circ
h\;d\mu\right)\right)d \lambda\right)\! = g^{1}\!\left(\int_Y
g\!\left(f^{1}\!\left(\int_X f \circ h\;d\lambda\right)\right)d \mu\right)$$
is satisfied by every $\mathscr{L} \otimes \mathscr{M}$measurable simple
function $h: X \times Y \to I$ if and only if $f=c g$ for some $c \in
\mathbf{R}^+$ (it is easy to see that the equation is well posed). An
analogous, but essentially different, result, with $f$ and $g$ replaced by
continuous injections $I \to \mathbf R$ and $\lambda(X)=\mu(Y)=1$, was recently
obtained in [Indag. Math. 27 (2016), 945953].

Let $H$ be a commutative semigroup with unit element such that every nonunit
can be written as a finite product of irreducible elements (atoms). For every
$k \in \mathbb N$, let $\mathscr U_k (H)$ denote the set of all $\ell \in
\mathbb N$ with the property that there are atoms $u_1, \ldots, u_k, v_1,
\ldots, v_{\ell}$ such that $u_1 \cdot \ldots \cdot u_k = v_1 \cdot \ldots
\cdot v_{\ell}$ (thus, $\mathscr U_k (H)$ is the union of all sets of lengths
containing $k$).
The Structure Theorem for Unions states that, for all sufficiently large $k$,
the sets $\mathscr U_k (H)$ are almost arithmetical progressions with the same
difference and global bound. We present a new approach to this result in the
framework of arithmetic combinatorics, by deriving, for suitably defined
families of subsets of the nonnegative integers, a characterization of when
the Structure Theorem holds.
This abstract approach allows us to verify, for the first time, the Structure
Theorem for a variety of possibly noncancellative semigroups, including
semigroups of (not necessarily invertible) ideals and semigroups of modules.
Furthermore, we provide the very first example of a semigroup (actually, a
locally tame Krull monoid) that does not satisfy the Structure Theorem.

Let $\mathcal{P}({\bf N})$ be the power set of $\bf N$. An upper density (on
$\bf N$) is a non\decreasing and subadditive function $\mu^\ast:
\mathcal{P}({\bf N})\to\bf R$ such that $\mu^\ast({\bf N}) = 1$ and $\mu^\ast(k
\cdot X + h) = \frac{1}{k} \mu^\ast(X)$ for all $X \subseteq \bf N$ and $h,k
\in {\bf N}^+$, where $k \cdot X + h := \{kx + h: x \in X\}$.
The upper asymptotic, upper Banach, upper logarithmic, upper Buck, upper
P\'olya, and upper analytic densities are examples of upper densities.
We show that every upper density $\mu^\ast$ has the strong Darboux property,
and so does the associated lower density, where a function $f: \mathcal P({\bf
N}) \to \bf R$ is said to have the strong Darboux property if, whenever $X
\subseteq Y \subseteq \bf N$ and $a \in [f(X),f(Y)]$, there is a set $A$ such
that $X\subseteq A\subseteq Y$ and $f(A)=a$. In fact, we prove the above under
the assumption that the monotonicity of $\mu^\ast$ is relaxed to the weaker
condition that $\mu^\ast(X) \le 1$ for every $X \subseteq \bf N$.

Classical Gon\v{c}arov polynomials are polynomials which interpolate
derivatives. Delta Gon\v{c}arov polynomials are polynomials which interpolate
delta operators, e.g., forward and backward difference operators. We extend
fundamental aspects of the theory of classical bivariate Gon\v{c}arov
polynomials and univariate delta Gon\v{c}arov polynomials to the multivariate
setting using umbral calculus. After introducing systems of delta operators, we
define multivariate delta Gon\v{c}arov polynomials, show that the associated
interpolation problem is always solvable, and derive a generating function (an
Appell relation) for them. We show that systems of delta Gon\v{c}arov
polynomials on an interpolation grid $Z \subseteq \mathbb{R}^d$ are of binomial
type if and only if $Z = A\mathbb{N}^d$ for some $d\times d$ matrix $A$. This
motivates our definition of delta Abel polynomials to be exactly those delta
Gon\v{c}arov polynomials which are based on such a grid. Finally, compact
formulas for delta Abel polynomials in all dimensions are given for separable
systems of delta operators. This recovers a former result for classical
bivariate Abel polynomials and extends previous partial results for classical
trivariate Abel polynomials to all dimensions.

Let $f$ and $g$ be realvalued continuous injections defined on a nonempty
real interval $I$, and let $(X, \mathscr{L}, \lambda)$ and $(Y, \mathscr{M},
\mu)$ be probability spaces in each of which there is at least one measurable
set whose measure is strictly between $0$ and $1$.
We say that $(f,g)$ is a $(\lambda, \mu)$switch if, for every $\mathscr{L}
\otimes \mathscr{M}$measurable function $h: X \times Y \to \mathbf{R}$ for
which $h[X\times Y]$ is contained in a compact subset of $I$, it holds $$
f^{1}\!\left(\int_X f\!\left(g^{1}\!\left(\int_Y g \circ
h\;d\mu\right)\right)d \lambda\right)\! = g^{1}\!\left(\int_Y
g\!\left(f^{1}\!\left(\int_X f \circ h\;d\lambda\right)\right)d \mu\right)\!,
$$ where $f^{1}$ is the inverse of the corestriction of $f$ to $f[I]$, and
similarly for $g^{1}$.
We prove that this notion is welldefined, by establishing that the above
functional equation is wellposed (the equation can be interpreted as a
permutation of generalized means and raised as a problem in the theory of
decision making under uncertainty), and show that $(f,g)$ is a $(\lambda,
\mu)$switch if and only if $f = ag + b$ for some $a,b \in \mathbf R$, $a \ne
0$.

Let $\mathbb G = (G, +)$ be a group (either abelian or not). Given $X, Y
\subseteq G$, we denote by $\langle Y \rangle$ the subsemigroup of $\mathbb G$
generated by $Y$, and we set $$\gamma(Y) := \sup_{y_0 \in Y} \inf_{y_0 \ne y
\in Y} {\rm ord}(y  y_0)$$ if $Y \ge 2$ and $\gamma(Y) := Y$ otherwise. We
prove that if $\langle Y \rangle$ is commutative, $Y$ is nonempty, and $X+2Y
\neq X + Y + y$ for some $y \in Y$, then $$ X+Y \ge X+\min(\gamma(Y), Y 
1). $$ Actually, this is obtained from a more general result, which improves on
previous work of the author on sumsets in cancellative semigroups, and yields a
comprehensive generalization, and in some cases a considerable strengthening,
of various additive theorems, notably including the ChowlaPillai theorem (on
sumsets in finite cyclic groups) and the specialization to abelian groups of
the HamidouneShatrowsky theorem.

The primary goal of this paper is to abstract notions, results and
constructions from the theory of categories to the broader setting of plots.
Loosely speaking, a plot can be thought of as a nonassociative nonunital
category with a "relaxed" composition law: Besides categories, this includes as
a special case graphs and neocategories in the sense of Ehresmann, Gabriel's
quivers, Mitchell's semicategories, and composition graphs, precategories and
semicategories in the sense of Schr\"oder. Among other things, we formulate an
"identityfree" definition of isomorphisms, equivalences, and limits, for which
we introduce regular representations, punctors, $\mathcal M$connections, and
$\mathcal M$factorizations. Part of the material will be used in subsequent
work to lay the foundation for an abstract theory of "normed structures"
serving as a unifying framework for the development of fundamental aspects of
the theory of normed spaces, normed groups, etc., on the one hand, and measure
spaces, perhaps surprisingly, on the other.

We introduce the concept of a prenormed model of a particular kind of
finitary singlesorted firstorder theories, interpreted over a category with
finite products. These are referred to as prealgebraic theories, for the fact
that their signature comprises, together with arbitrary function symbols (of
finite ariety), only relation symbols whose interpretation, in any possible
model, is a reflexive and transitive binary relation, namely a preorder. The
result is an abstract approach to the very concept of norm and, consequently,
to the theory of normed structures.

We introduce the sequence of generalized Gon\v{c}arov polynomials, which is a
basis for the solutions to the Gon\v{c}arov interpolation problem with respect
to a delta operator. Explicitly, a generalized Gon\v{c}arov basis is a sequence
$(t_n(x))_{n \ge 0}$ of polynomials defined by the biorthogonality relation
$\varepsilon_{z_i}(\mathfrak d^{i}(t_n(x))) = n! \;\! \delta_{i,n}$ for all
$i,n \in \mathbf N$, where $\mathfrak d$ is a delta operator, $\mathcal Z =
(z_i)_{i \ge 0}$ a sequence of scalars, and $\varepsilon_{z_i}$ the evaluation
at $z_i$. We present algebraic and analytic properties of generalized
Gon\v{c}arov polynomials and show that such polynomial sequences provide a
natural algebraic tool for enumerating combinatorial structures with a linear
constraint on their order statistics.

Given an additively written abelian group $G$ and a set $X\subseteq G$, we
let $\mathscr{B}(X)$ denote the monoid of zerosum sequences over $X$ and
$\mathsf{D}(X)$ the Davenport constant of $\mathscr{B}(X)$, namely the supremum
of the positive integers $n$ for which there exists a sequence $x_1 \cdots x_n$
of $\mathscr{B}(X)$ such that $\sum_{i \in I} x_i \ne 0$ for each nonempty
proper subset $I$ of $\{1, \ldots, n\}$. In this paper, we mainly investigate
the case when $G$ is a power of $\mathbb{Z}$ and $X$ is a box (i.e., a product
of intervals of $G$). Some mixed sets (e.g., the product of a group by a box)
are studied too, and some inverse results are obtained.

We address the "sums of dilates problem" by looking for nontrivial lower
bounds on sumsets of the form $k \cdot X + l \cdot X$, where $k$ and $l$ are
nonzero integers and $X$ is a subset of a possibly nonabelian group $G$
(written additively). In particular, we investigate the extension of some
results so far known only for the integers to the context of torsionfree or
linearly orderable groups, either abelian or not.

Let $\mathbb{A} = (A, +)$ be a (possibly noncommutative) semigroup. For $Z
\subseteq A$ we define $Z^\times := Z \cap \mathbb A^\times$, where $\mathbb
A^\times$ is the set of the units of $\mathbb{A}$, and $$\gamma(Z) := \sup_{z_0
\in Z^\times} \inf_{z_0 \ne z \in Z} {\rm ord}(z  z_0).$$ The paper
investigates some properties of $\gamma(\cdot)$ and shows the following
extension of the CauchyDavenport theorem: If $\mathbb A$ is cancellative and
$X, Y \subseteq A$, then $$X+Y \ge \min(\gamma(X+Y),X + Y  1).$$ This
implies a generalization of Kemperman's inequality for torsionfree groups and
strengthens another extension of the CauchyDavenport theorem, where
$\mathbb{A}$ is a group and $\gamma(X+Y)$ in the above is replaced by the
infimum of $S$ as $S$ ranges over the nontrivial subgroups of $\mathbb{A}$
(HamidouneK\'arolyi theorem).

Given an integer $n \ge 3$, let $u_1, \ldots, u_n$ be pairwise coprime
integers $\ge 2$, $\mathcal D$ a family of nonempty proper subsets of $\{1,
\ldots, n\}$ with "enough" elements, and $\varepsilon$ a function $ \mathcal D
\to \{\pm 1\}$. Does there exist at least one prime $q$ such that $q$ divides
$\prod_{i \in I} u_i  \varepsilon(I)$ for some $I \in \mathcal D$, but it does
not divide $u_1 \cdots u_n$? We answer this question in the positive when the
$u_i$ are prime powers and $\varepsilon$ and $\mathcal D$ are subjected to
certain restrictions. We use the result to prove that, if $\varepsilon_0 \in
\{\pm 1\}$ and $A$ is a set of three or more primes that contains all prime
divisors of any number of the form $\prod_{p \in B} p  \varepsilon_0$ for
which $B$ is a finite nonempty proper subset of $A$, then $A$ contains all the
primes.

Let $\mathbb{A} = (A, \cdot)$ be a semigroup. We generalize some recent
results by G. A. Freiman, M. Herzog and coauthors on the structure theory of
set addition from the context of linearly orderable groups to linearly
orderable semigroups, where we say that $\mathbb{A}$ is linearly orderable if
there exists a total order $\le$ on $A$ such that $xz < yz$ and $zx < zy$ for
all $x,y,z \in A$ with $x < y$. In particular, we find that if $S$ is a finite
subset of $A$ generating a nonabelian subsemigroup of $\mathbb{A}$, then
$S^2 \ge 3S2$. On the road to this goal, we also prove a number of
subsidiary results, and most notably that for $S$ a finite subset of $A$ the
commutator and the normalizer of $S$ are equal to each other.

We generalize the Davenport transform and use it to prove that, for a
(possibly noncommutative) cancellative semigroup $\mathbb A = (A, +)$ and
nonempty subsets $X,Y$ of $A$ such that the subsemigroup generated by $Y$ is
commutative, we have $X + Y \ge \min(\omega(Y), X + Y  1)$, where
$\omega(Y) := \sup_{y_0 \in Y \cap \mathbb A^{\times}} \inf_{y \in Y \setminus
\{y_0\}} <y  y_0>$. This carries over the CauchyDavenport theorem to the
broader setting of semigroups, and it implies, in particular, an extension of
I. Chowla's and S.S. Pillai's theorems for cyclic groups and a notable
strengthening of another generalization of the same CauchyDavenport theorem to
commutative groups, where $\omega(Y)$ in the above is replaced by the minimal
order of the nontrivial subgroups of $\mathbb A$.

We determine all triples $(a,b,n)$ of positive integers such that $a$ and $b$
are relatively prime and $n^k$ divides $a^n + b^n$ (respectively, $a^n  b^n$),
when $k$ is the maximum of $a$ and $b$ (in fact, we answer a slightly more
general question). As a byproduct, it is found that, for $m, n \in \mathbb
N^+$ with $n \ge 2$, $n^m$ divides $ m^n + 1$ if and only if $(m,n)=(2,3)$ or
$(1,2)$, which generalizes problems from the 1990 and 1999 editions of the
International Mathematical Olympiad. The results are related to a conjecture by
K. Gy\H{o}ry and C. Smyth on the finiteness of the sets $R_k^\pm(a,b) := \{n
\in \mathbb N^+: n^k \mid a^n \pm b^n\}$, when $a,b,k$ are fixed integers with
$k \ge 3$, $\gcd(a,b) = 1$ and $ab \ge 2$.