the Human Gene Nomenclature Consortium (HGNC) has always tried to bring some order into the caos that gene nomenclature was suffering for years, and what they've came out with is the concept of "approved name" and "alias". what they've always tried is to select a gene name as the approved one, leaving all the synonyms they found reported for that gene as aliases.

although HGNC gene names are commonly considered as the gold standard, and although lots of efforts have been made in order to standarize this issue, not every entry on the main gene informative repositories (UCSC, Ensembl, NCBI's Entrez, ...) follow this rule. for that reason you may still find some gene names inside UCSC that may refer to the same single HGNC approved gene name.

by UCSC "gene" do you mean the names from the table knownGene ? It is a database of transcripts, that is why they map to one HGNC 'Gene' (Many-to-one).