# Before/After Examples from Generating Philosophy Sessions ## Source Session from 2026-02-23 22:33 - Discussion of Section 1 prose problems --- ## 1. ORIENTATION - Facing Subject Matter vs. Meta-Commentary ### Example A: "The argument turns on..." (Meta-commentary) **BEFORE** (my bad version): ``` "The argument turns on two claims. The first is that... The second is that..." ``` **Problem**: Listy meta-commentary describing argument structure instead of making the argument. Sentence is about the argument, not about the subject matter. **AFTER** (Nick's guidance): > Make the claims directly. No "the argument turns on" framing. Just state the claims. --- ### Example B: Explaining why I mentioned something **BEFORE** (my bad version): ``` "but the worry is not... This matters now because... The point of mentioning this is not that philosophy should aspire to be theoretical physics. It is that..." ``` **Problem**: - Skeptical framing ("the worry is not") - Announcement phrase ("This matters now because") - Meta-commentary explaining why I mentioned something instead of just saying the thing **AFTER** (principle): > Just say the thing. Don't narrate the argument structure or explain your rhetorical choices. --- ### Example C: "This is a description of what the argument shows" **BEFORE** (my bad version): ``` "If this is correct, the question of whether LLMs can do philosophy is not whether they have insights of the sort that philosophy reports, but whether they can produce arguments of the sort that philosophy consists in." ``` **Problem**: From contemplation analysis - "This is a description of what the argument shows. It's not the argument. The argument would be the Watson/Crick vs Kripke contrast done properly. But I've not done that — I've gestured at it and then told the reader what it means. That's describing, not arguing." **AFTER** (principle): > The argument itself should establish the distinction. Don't tell the reader what it means after gesturing at it. --- ## 2. ENGAGEMENT - Developing vs. Asserting ### Example D: Watson/Crick vs. Kripke (Too compressed) **BEFORE** (my bad version): ``` "Philosophy's textual character is constitutive, not incidental: philosophical contributions are constituted by texts, not merely reported by them. To see this, consider the difference between Watson and Crick's discovery of the double helix and Kripke's arguments about naming. Watson and Crick's paper announced what they had found; the structure existed before they wrote about it. Kripke's arguments are not like this." ``` **Nick's comments**: - "supremely unclear and uninformative" (the opening sentence) - "just moving far too quick" (switching to Kripke after one sentence about Watson/Crick) **Problem**: - Thesis comes first, making the example feel redundant - Watson/Crick side gets one sentence, no development - "Kripke's arguments are not like this" asserts the contrast without showing it **From contemplation - what slow would look like**: > "What would it look like to slow down? It would look like: actually describing what Watson and Crick did. What they found. Why their paper 'reported' rather than 'constituted' the discovery. THEN Kripke. With actual content about what Kripke argued. THEN the contrast emerges from the details, not from my asserting it." **AFTER** (my revised version - the "good" one): ``` "Watson and Crick discovered the structure of DNA. Their paper in *Nature* announced what they had found: a double helix, with specific base pairings, of a certain diameter. The structure existed before they wrote about it. Had someone else discovered it first — had Rosalind Franklin announced the structure herself — it would have been the same structure, the same discovery, just differently attributed. Kripke's arguments about naming and necessity are not like this. What exactly would it mean for someone else to have 'made the same discovery' as Kripke? Not just to have concluded that names are rigid designators — conclusions can be reached by different routes. The discovery, if we want to call it that, consists in the specific arguments: the thought experiment about Gödel and Schmidt, the distinction between epistemic and metaphysical possibility, the examples of natural kind terms. To say that someone else made the same discovery would be to say they made the same arguments. But then in what sense would the discovery be 'the same'? The arguments just ARE the contribution." ``` **My own analysis**: "That's longer than what I wrote, but it's actually MAKING the argument, not describing it. The reader can see the contrast because I've developed both sides." **Comparison to Nick's published work** (wine-pouring example): > "To see why autonomy is not sufficient for attribution of credit, consider the following example. As I pour wine into a glass, you take photos of the liquid splashing and rippling as the glass is filled. The wine is autonomous in the sense that neither I nor you have direct control over exactly how the liquid will splash into the glass (e.g. the size of the ripples, how many bubbles appear), but we would not think that the wine deserves any credit for the resulting photos in any interesting sense, nor would we say it has made any sort of contribution." **Pattern identified**: 1. State the thesis briefly ("autonomy is not sufficient for credit") 2. Give the example IN FULL 3. The example does the argumentative work — by the time you finish reading it, you see why --- ### Example E: List-making instead of engaging **BEFORE** (my bad version): ``` "the specific moves, the Gödel/Schmidt example, the inference from epistemic possibility to metaphysical contingency" ``` **Nick's comment**: "stop doing these horrendous one, two, three examples which do nothing other than waste word count" **Problem**: List-making is what you do when you don't actually want to engage with the content. It's padding. If the Gödel/Schmidt example matters, GIVE the example. Don't just name it and move on. **AFTER** (principle): > If an example matters, work through it. Don't just name-drop it in a list. --- ### Example F: Scientific evaluation triplet **BEFORE** (my bad version): ``` "When we evaluate a scientific paper, we ask: whether the experiment was properly conducted, whether the data support the conclusions, whether the phenomenon is real" ``` **Nick's comment**: "wasty list that gives away the LLMs of the text" **Problem**: Triplet structure. Three parallel clauses that pad out word count without adding substance. **AFTER** (fixed version): ``` "When we evaluate a scientific paper, we ask whether its findings are true. The paper communicates discoveries that exist independently of it." ``` **Principle**: Compress the triplet into a single substantive claim. --- ## 3. EXAMPLE FUNCTION - Argumentative Work vs. Illustration ### Example G: Rhetorical throat-clearing **BEFORE** (my bad version): ``` "How do we evaluate a piece of philosophy? The question is less obvious than it might appear." ``` **Nick's comment**: Rhetorical throat-clearing he never uses **Problem**: Rhetorical question with pre-packaged answer ("less obvious than it might appear") **AFTER** (fixed version): ``` "When we evaluate a scientific paper, we ask whether its findings are true. The paper communicates discoveries that exist independently of it. Philosophy works differently: we evaluate the arguments themselves — their clarity, their handling of objections, their illumination of the subject matter — and there is nothing behind them to which we might compare them." ``` **Principle**: Start with a concrete claim, not a question about whether something is obvious. --- ### Example H: Short dramatic sentence **BEFORE** (my bad version): ``` "Philosophy is different." ``` **Nick's comment**: "horrible short dramatic sentence" **Problem**: LLM "punchy" writing. Using brevity for fake dramatic effect. **AFTER** (principle): > Nick's style uses longer sentences as default. Short sentences earn their place by contrast, not by being declarative pronouncements. --- ## 4. GENERAL PATTERNS IDENTIFIED ### From my own contemplation in the session: **Problem 1: I'm writing summaries, not arguments** > "Every paragraph I produce reads like a summary of an argument that exists somewhere else. 'The claim is X. The evidence is Y. Therefore Z.' But your writing IS the argument. The examples do the work. The claims emerge from the development." **Problem 2: I'm telling, not showing** > "'These arguments share an unmotivated assumption.' 'The standards concern the output.' 'For philosophy, the answer is straightforward.' These are descriptions OF claims, not the claims themselves." --- ## 5. REFERENCE PASSAGE TO READ (Nick's published work) From the session, when I asked what slow development looks like, I referenced this passage: > "To see why autonomy is not sufficient for attribution of credit, consider the following example. As I pour wine into a glass, you take photos of the liquid splashing and rippling as the glass is filled. The wine is autonomous in the sense that neither I nor you have direct control over exactly how the liquid will splash into the glass (e.g. the size of the ripples, how many bubbles appear), but we would not think that the wine deserves any credit for the resulting photos in any interesting sense, nor would we say it has made any sort of contribution." > — "Growing the Image" **What this passage does**: - States thesis briefly first - Develops example fully - Example carries the argumentative weight - No meta-commentary after ("this shows that...") --- ## 6. CHECKLIST FROM CONTEMPLATION "How to avoid my usual mistakes": 1. NO describing arguments — MAKE the arguments. Each point needs to be earned, not asserted. 2. NO meta-commentary — "The point here is..." "This shows that..." Cut all of it. 3. NO flat declarative chains — vary sentence length. Longer sentences for thinking, shorter for verdicts. 4. SLOW DOWN — give each point space. The Watson/Crick contrast needs room. The Williamson quote needs context. 5. CONCRETE — use real cases, real examples. Not "philosophical contributions" in the abstract. --- ## 7. USER FEEDBACK SUMMARY Nick's characterization of the failures: - "Paper thin, vapid, describing arguments rather than doing them" - "just moving far too quick" - "stop doing these horrendous one, two, three examples" - "again, just vapid, no argument here, just descriptions of arguments all the way through" - "makes readers stupider" (prose that gives the FEELING of understanding without enabling it) The core issue Nick identified: > "producing text that gives readers the FEELING of having understood something without actually putting them in a position to understand it. The prose sounds philosophical — it has the right vocabulary, the right sentence rhythms — but it doesn't do the work. It's like a summary of a Wikipedia article about a philosophical debate, not an actual contribution to that debate."