On 1-cycles and the finite dimensionality of homology 4-manifolds (Q1208255)

The cell-like maps on manifolds of dimension \(\leq 3\) do not raise dimension, while on manifolds of dimension \(\geq 5\) they can raise dimension. The authors attack the problem of raising dimension by cell- like maps on 4-manifolds as the remaining unsettled case. They have obtained the following result in this direction: Let \(M^ 4\) be a topological 4-manifold and \(f: M^ 4 \to X\) be a proper cell-like onto map. Then \(\dim X < \infty\) if and only if for some \(n \geq 3\), \(X\) has disjoint Pontrjagin \(n\)-tuples property, denoted \(dd_ n\). Property \(dd_ n\) means the following: For every \(\varepsilon > 0\) and every collection of maps \(f_ 1,\dots,f_ n: \mathbb{D}^ 2 \to X\) of the Pontrjagin disc \(\mathbb{D}^ 2\) into \(X\), there exist maps \(g_ 1,\dots,g_ n: \mathbb{D}^ 2 \to X\) such that (i) for every \(i\), \(d(f_ i,g_ i) < \varepsilon\), and (ii) \(\bigcap^ n_{i=1}g_ i(\mathbb{D}^ 2) = \emptyset\). This result is the corollary of a theorem which gives the criterion of finite dimensionality of \(\mathbb{Z}\)-homology 4-manifolds. This theorem reads as follows: Let \(X\) be a \(\mathbb{Z}\)-homology 4-manifold. Then \(\dim X < \infty\) (equivalently \(\dim X = 4\)) if and only if for some \(n \geq 3\) \(X\) has \(dd_ n\). The proof of finite dimensionality of \(X\) involves different results and it contains statements on representations of 1-cycles which are of interest in their own, and in final it is based on presenting \(X\) as the union of two finite dimensional pieces. It is also pointed out that \(dd_ 3\) is more general than \(DD_ 3\) (disjoint triples property of R. J. Daverman). Replacing the Pontrjagin disc \(\mathbb{D}^ 2\) in the definition of \(dd_ 3\) by the disc \(D^ 2\) one has the definition of \(DD_ 3\).