From a statistical point of view, correlation is usually limited to a specific form of association between two interval-level variables. But if a pair of categorical variables both use the same set of categories (i.e., you have a "square" table), then there are any number on non-parametric coefficients you can use to assess the association in such data.
In general, association is a linguistic term used in writing, while correlation is a statistical term describes the significant and degree of that association. Regards.
Correlation refers to the statistical relationship between two variables. It measures the degree to which changes in one variable correspond to changes in another variable. Correlation can be positive (both variables increase or decrease together), negative (one variable increases while the other decreases), or zero (no relationship).
Association, on the other hand, refers to a general relationship or connection between two variables. It does not necessarily imply a statistical relationship or indicate the strength or direction of the relationship. Association can be based on various factors such as causality, similarity, or coincidence.
Interesting question and a common source of confusion.
I think all correlations are measures of association but not all measures of association are correlations. Both are numerical expressions of strength and direction of statistical association/'co-relation', but have different ranges and null values.
Commonly used correlations are expressed as coefficients that range from -1.0 to 1.0 with a null value of 0. They can be calculated for interval-ratio (Pearson) rank/ordinal (Spearman, gamma, tau etc.) binary (phi) binary x continuous (point biserial) data, etc.
Commonly used ratio measures of association range from 0 to infinity with a null value of 1.0 and can be calculated between groups for relative odds of exposure (Odds Ratio), probabilities of outcome (Risk Ratio), time to event (Hazard Ratio) or events per person-time at risk (Incidence Rate Ratio).