This paper positions Prompt Engineering within the broader context of Discourse Analysis (DA) and Human-Computer Interaction (HCI), highlighting the transformative potential of Speech Act Theory in the age of AI-driven dialogue systems. Traditionally treated as a technical exercise involving precise command inputs, prompt crafting is reconceptualized as a dialogical and intentional communicative practice that derives from human language. Drawing on foundational work by Austin et al. the paper argues that interactions with large language models can be framed as speech acts which are expressions shaped by intention, social function, and a turn-taking structure. Aimed at non-expert users seeking more effective ways to converse with chatbots, the paper introduces a short pattern language that is derived from illocutionary speech acts. It identifies six recurring types of prompting that represent a reflection of natural conversational behavior, providing a practical framework for designing meaningful, context-aware, and goal-oriented exchanges between human and machine. By bridging computational logic with human discourse conventions, the paper advocates for a shift from isolated directive one-shot prompts to a multi-turn, purposeful dialogue to enhance interpretability, relevance, and user agency in AI-mediated communication.