• AThe LALR(1) parser for a grammar G cannot have reduce-reduce conflict if the LR(1) parser for G does not have reduce-reduce conflict.
• BSymbol table is accessed only during the lexical analysis phase.
• CData flow analysis is necessary for run-time memory management.
• LR(1) parsing is sufficient for deterministic context-free languages.

(A) is FALSE: merging LR(1) states with identical cores to form LALR(1) states CAN introduce NEW reduce-reduce conflicts that weren't present in the canonical LR(1) automaton — a well-known LALR limitation. (B) is FALSE: the symbol table is accessed throughout compilation (semantic analysis, code generation), not just lexical analysis. (C) is FALSE: data flow analysis is used for optimization, not for run-time memory management. (D) is TRUE: it's a foundational theorem that the class of languages parsable by LR(1) grammars is EXACTLY the class of deterministic context-free languages — LR(1) parsing is both necessary and sufficient.