AgentSkillsCN

dissertation-prospectus

每当用户需要帮助撰写、起草、修改或梳理人类学研究的论文提纲、论文提案、资格考试提案、升级文件、转学文件,或田野调查许可提案时,均可使用此技能。触发条件包括:任何提及“论文提纲”、“论文提案”、“资格考试”、“QE”、“升级提案”、“身份转换”、“身份确认”、“田野调查提案”或“田野调查许可”等短语,且这些短语均与人类学或民族志研究相关;请求对论文提案的任一章节进行结构化、起草或修改(问题陈述、研究问题、理论定位、文献综述、方法、伦理、时间表、预算);询问委员会的期望是什么,或如何为论文提纲答辩或升级口试做准备。当用户说“我需要写我的论文提纲”、“我正在准备资格考试”、“我该如何结构化论文提案”、“我的论文提纲应该包含什么”时,也可使用此技能。本技能涵盖美国的论文提纲(伯克利、哈佛等项目)、英国的升级/转学/田野调查提案(LSE、剑桥、牛津),以及兼具论文提纲与资助申请双重用途的提案。切勿用于无委员会听众的独立资助提案(请使用资助提案技能)、一般性的学术论文写作(请使用学术论文技能),或在没有论文提纲背景的情况下进行研究问题开发(请使用研究问题技能)。

SKILL.md
--- frontmatter
name: dissertation-prospectus
description: >
  Use this skill whenever a user needs help writing, drafting, revising, or
  structuring a dissertation prospectus, dissertation proposal, qualifying exam
  proposal, upgrade document, transfer document, or fieldwork clearance proposal
  for anthropological research. Triggers include: any mention of "prospectus,"
  "dissertation proposal," "qualifying exam," "QE," "upgrade proposal,"
  "transfer of status," "confirmation of status," "fieldwork proposal," or
  "fieldwork clearance" in the context of anthropology or ethnographic research;
  requests to structure, draft, or revise any section of a dissertation proposal
  (problem statement, research questions, theoretical positioning, literature
  review, methods, ethics, timeline, budget); questions about what a committee
  expects or how to prepare for a prospectus defense or upgrade viva. Also use
  when the user says "I need to write my prospectus," "I'm preparing for my
  qualifying exam," "how do I structure a dissertation proposal," or "what
  should my prospectus include." Covers US prospectuses (Berkeley, Harvard, and
  other programs), UK upgrade/transfer/fieldwork proposals (LSE, Cambridge,
  Oxford), and dual-purpose prospectuses that also serve as grant applications.
  Do NOT use for standalone grant proposals without a committee audience (use
  grant-proposal skill), general academic paper writing (use academic-paper
  skill), or research question development without a prospectus context (use
  research-question skill).

Dissertation Prospectus Writing

Write dissertation prospectuses and proposals for anthropological research that satisfy both committee expectations (intellectual coherence, feasibility, scholarly preparation) and — when dual-purpose — funder requirements (compliance, risk mitigation, credible execution). The strongest prospectuses are modular: adaptable into a committee document, a fundable grant narrative, and an ethics dossier without full rewrites.

Quick Reference

TaskReference
Full prospectus guidance (sections, length norms, evaluation criteria, examples)Read references/prospectus-guide.md
NSF DDRIG-specific requirements (if dual-purpose)Load from grant-proposal skill: references/nsf-cultural-anthro.md
Wenner-Gren-specific requirements (if dual-purpose)Load from grant-proposal skill: references/wenner-gren.md
Fulbright-specific requirements (if dual-purpose)Load from grant-proposal skill: references/fulbright.md

Workflow

Step 1: Identify What the User Needs

Determine the entry point:

  • Writing from scratch. The user has a topic and needs to build a full prospectus. Load the full reference file and work through sections in order.
  • Revising an existing draft. The user has a draft and wants feedback or help improving specific sections. Identify which sections need work and load the relevant guidance from the reference file.
  • Adapting format. The user has a prospectus for one context and needs to adapt it for another (e.g., committee prospectus → NSF DDRIG application). Load both the prospectus reference and the relevant funder reference.

Step 2: Gather Context

Before generating any content, ask for these required inputs:

  1. Program and milestone. Which institution? Which milestone (US qualifying exam, UK upgrade/transfer, UK fieldwork clearance, other)? This determines length norms, required components, and evaluation criteria.
  2. Research topic and site(s). What is the project about and where will fieldwork occur?
  3. Epistemic stance. Which theoretical orientation(s) does the researcher work within? Ask which is primary and which are secondary.
  4. Target funder (if dual-purpose). Is this prospectus also serving as the basis for an NSF DDRIG, Wenner-Gren, Fulbright, or other application?
  5. Stage of development. Starting from scratch, refining a draft, or adapting to a different format?

Helpful but not required: theoretical framework, preliminary fieldwork or pilot data, language competencies, committee composition, timeline constraints.

Step 3: Load Appropriate Reference

Always load references/prospectus-guide.md for the full section-by-section guidance, length norms, evaluation criteria, and examples.

If the user is writing a dual-purpose document, also load the relevant funder reference from the grant-proposal skill to ensure the prospectus satisfies both audiences simultaneously.

Step 4: Generate Output

Follow the section structure and proportional allocations in the reference file. Key principles:

  • Modular design. Each section should work both for the committee and for potential funder adaptation.
  • Theory as leverage. Theoretical positioning shows what conceptual work becomes possible, not just allegiance to a framework.
  • Methods as inferential design. Specify evidence, justify the site-method pairing, and show how analysis proceeds.
  • Ethics as first-class content. Address consent modality, community obligations, and data protection substantively — not as boilerplate.
  • Contingency planning. Include a Plan B methodology.
  • Match length and format norms. US prospectuses range from 8-10 pages (Berkeley) to 25-30 pages (Harvard). UK proposals range from 7,000 to 20,000 words depending on milestone.

Step 5: Quality Check

Before presenting output, verify:

  • Research questions are fieldwork-operational AND theoretically motivated
  • Theoretical positioning shows leverage, not just allegiance
  • Literature review is selective, problem-centered, and multi-tradition
  • Methods section specifies evidence, justifies site-method pairing, and includes analysis plan
  • Ethics section addresses consent modality, community obligations, and data protection substantively
  • Timeline includes pre-fieldwork workstreams
  • Budget is itemized and connected to the evidence plan
  • Contingency plan is present
  • Tone is confident without overclaiming
  • Output matches the user's program length and format requirements
  • Epistemic stance is reflected consistently throughout
  • If dual-purpose: all funder-specific requirements are met

Parameters

  • Epistemic stance: All 42 stances are relevant. Most dissertations combine a primary and one or more secondary stances.
  • Genre/audience: Committee prospectus, dual-purpose prospectus + grant, UK upgrade/transfer, UK fieldwork clearance.
  • Compression: 8-10 pages (short US), 20-30 pages (long US), 7,000-20,000 words (UK, varies by milestone).
  • Risk posture: Low-risk, vulnerable populations, high-surveillance, politically sensitive.
  • Formality register: Disciplinary (default for committee audiences).
  • Field configuration: Single site, multi-sited, digital, archival, hybrid.

Guardrails

  • Do not generate content without knowing the program and milestone. Length norms, required components, and evaluation criteria differ significantly between US and UK programs and between institutions. Ask first.
  • Do not produce ethics sections that are boilerplate. Ethics must address the specific consent modality, community obligations, and data protection concerns relevant to the project. If the risk posture is elevated, ethics protections must match.
  • Do not omit contingency planning. All prospectuses should include a Plan B methodology, even when not formally required.
  • Flag dual-purpose tension. When a prospectus serves both committee and funder audiences, flag any points where their expectations diverge (e.g., NSF requires falsifiable questions; committees may accept more interpretive framing).
  • Do not overclaim. Epistemic humility about what the research will and will not be able to show is a mark of sophistication.

Common Failure Modes

Failure modePrevention
Overbreadth — too many questions, none convincingLimit to 2-4 tightly articulated questions
Ethics as afterthought — perfunctory sectionMake ethics a workflow step throughout
Vague analysis plan — "I will use thematic analysis"Specify coding approach, triangulation, what constitutes a claim
Methods as menu — lists techniques without logicFrame as inferential design generating specific evidence
No contingency — assumes everything goes to planInclude Plan B with alternative methods and claim limitations
Answers already known — telegraphs conclusionsFrame genuinely open questions; acknowledge uncertainty
Generic literature review — lists without argumentMake the review an argument about where this project intervenes
Mission mismatch — description without theory contributionConnect every methods choice to a theoretical claim

Examples

Example 1: US sociocultural prospectus (Berkeley-style)

Input: "I'm writing my prospectus for my QE at Berkeley. My project is about how Senegalese migrants in Paris use WhatsApp groups to maintain translocal kinship networks. I work within an interpretivist framework drawing on practice theory and digital anthropology."

Output approach:

  • Load references/prospectus-guide.md
  • Set length to 8-10 pages (Berkeley tight format)
  • Set epistemic stance to interpretivist + practice theory
  • Set field configuration to multi-sited (Paris + digital + Senegal)
  • Methods: participant observation in WhatsApp groups, life history interviews, multi-sited fieldwork logic
  • Analysis: iterative coding with practice-theoretical categories

Example 2: UK fieldwork proposal (Cambridge-style)

Input: "I need to write my fieldwork proposal for Cambridge clearance. My research examines Indigenous water governance in Bolivia's Altiplano. I work within decolonial and political ecology traditions."

Output approach:

  • Load references/prospectus-guide.md
  • Set length to 7,000 words
  • Set epistemic stance to decolonial + political ecology
  • Set risk posture to politically sensitive + Indigenous communities
  • Methods: foreground community authority, reciprocity, Indigenous knowledge sovereignty; include CARE Principles for data management
  • Flag: separate risk assessment form also needed

Example 3: Dual-purpose prospectus + NSF DDRIG

Input: "I'm writing my dissertation prospectus and I also want to adapt it for my NSF DDRIG application. My project studies algorithmic hiring systems in tech companies from an STS perspective."

Output approach:

  • Load references/prospectus-guide.md AND nsf-cultural-anthro.md
  • Generate committee prospectus first, then flag NSF-specific adaptations
  • NSF requires labeled intellectual merit and broader impacts sections, falsifiable questions, explicit analysis plan, data management plan
  • Frame STS questions with falsifiable stakes for NSF audience