Why did Searle think that manipulating symbols does not prove understanding?

John Searle argued that manipulating symbols does not prove understanding because such operations rely solely on the syntactic (formal) properties of symbols rather than their semantic (meaning-based) properties.1

In his Chinese room argument, Searle described a person who uses a manual to transform Chinese character inputs into appropriate outputs. Although this process allows the person to simulate the behavior of a Chinese speaker—potentially passing a Turing test—the person remains entirely ignorant of what the characters actually mean.231

Searle concluded that because the person (or a computer) is merely following rules to manipulate symbols without grasping their significance, the process cannot be equated with genuine understanding. He maintained that strong artificial intelligence is deficient because it reduces mental processes to computation, failing to account for the semantic properties essential to human thought. While functionalists argue that the system as a whole possesses understanding, Searle insists that even if an individual memorized the entire system, they would still lack any comprehension of the symbols they manipulate.213