Most human affordances and the sequences that describe them are concrete: real objects and events to be perceived and tangible behaviors to follow. Human language is also populated with abstract nouns and verbs which lack physical correlates, like freedom, imagination, and loyalty. How can a system of sequences that evolved to describe real-world affordances also incorporate such abstractions? Looking at the cell can provide clues. There are many layers of regulation, all yielding more complex metabolic affordances. For example, allosterism lets interactors have their functions configured by external signals. Interactors in the cell are grouped into blue-collar and white-collar enzymes, the former governing actual chemical reactions and the latter governing the behavior of other enzymes. Functional differences across tissues arise when the blue-collar enzymatic workforce is organized into a control hierarchy by white-collar sequences and their interactors. The most complex control hierarchies are gene regulatory networks, which guide real-time metabolism and behavior, and the build-out of the body plan during development. The sequences that govern gene regulatory networks operate at a high level of abstraction, many layers away from the nuts and bolts of metabolism and development. Like abstract nouns in language, they constitute a grammar of abstraction.