North Slope and Base Camp: a boutique bank’s philosophical system and its limits of language北坡与营地：一家精品投行的哲学系统及其语言边界

1. Introduction: The Burden of Proof a Name Brings

If the name is a measure, it must allow others to measure you with it.

An institution's name usually does no more than identify it. Once a name invokes a specific intellectual source, it turns round and demands evidence from the institution.

The name Gengxin belongs to the second case. In the firm's own published words: "Wittgenstein wrote the Tractatus Logico-Philosophicus in 1920 and 1921. In the sexagenary calendar those two years are Gengshen and Xinyou; take one character from each and you have Gengxin. The English name is Glacier, close to Gengxin in sound. That is the whole provenance." (「维特根斯坦写《逻辑哲学论》，是 1920 到 1921 年；这两年的干支纪年分别是庚申与辛酉，各取一字，就是庚辛。英文名是冰川（Glacier），与庚辛音近。来历就这么多。」) [10]

Note the last four characters: "That is the whole provenance" (「来历就这么多」). This is a self-imposed limit. It explicitly declines to expand the naming into a claim of intellectual lineage. In the same text the firm offers a different statement, weaker and more testable: "The name is not a label. It is a measure." (「这个名字不是一个标签，是一把尺子。」) [10]

The difference between a measure and a label is the starting point of everything this paper does. A label only needs to be displayed. A measure must allow others to measure you with it. If an institution declares that it is named after the years in which a philosopher wrote, it owes at least one answer: which rule in its method is done that way because of the book, and which merely happens to sound like it.

This paper is here to settle that account. It does not defend the naming, nor does it assume the naming holds.

One gap in the source text must be declared first. Before writing, this paper searched the full body text of the Glacier website home page and the Glacier Institute page word by word and did not find the term "base camp" (「大本营」). The closest wording on the website is "camp" (「营地」) — "Arrival is not the end; returning to camp is" (「抵达不是终点，回到营地才是」) [11] — together with the "fact base" (「事实底座」) that serves as the bottom component of the system [9]. This paper therefore uses "camp", which exists on the website, in its title and throughout, and not the unverified "base camp". This is not a stylistic choice. It is a discipline of evidence: a paper arguing "if you cannot say it clearly, do not write it" cannot begin by breaking that rule itself.

---

2. The Question: Why an Investment Bank Needs a "Philosophical System" Rather Than a Set of Values

Values cannot be falsified, so they bind no one; axioms can be falsified, and that is why they govern action.

In business writing, "philosophical system" is usually a polite name for a statement of values. To judge whether Glacier uses the term the same way, one must first look at what work it does.

Glacier states its founding intent as follows: "Glacier Capital's founding intent is to contribute to and refine the philosophical system of the boutique investment bank." (「庚辛的初心，是贡献和完善精品投行的哲学系统。」) [9] What follows immediately is not a vision but two judgments the firm calls "axioms":

"First: the moat of a hard-tech company grows in its labs and on its production lines, not in its pitch deck. Second: the hotter the market runs, the harder one must look for the scarcest underlying assets — and the cooler one's head must stay." [9] (「第一条：硬科技的壁垒长在实验室和产线上，不长在商业计划书里。第二条：越是市场狂热时，越要寻找最稀缺的底层资产，内心也越要保持绝对冷静。」)

The form of these two deserves attention. They are falsifiable empirical propositions, not value preferences. "The moat grows in its labs and on its production lines" (「壁垒长在实验室和产线上」) is a claim about where technical barriers are located. If the barriers of a given generation of technology grow mainly in distribution channels or data flywheels, the axiom fails in that field.

This is what separates a "philosophical system" from "values". Values cannot be falsified ("we value integrity" cannot be overturned by data), and so they carry no constraint. Axioms can be falsified, and so they constrain every subsequent action. From the two axioms Glacier derives "We take the industry's north slope" (「坚定选择产业的『北坡』」) [9], and publishes the derivation as "the firm's position follows on its own" (「选择不言自明」).

The chain of inference is public, the premises are refutable, and the conclusion can be held to account. In this respect "philosophical system" is not a polite name here but a structural commitment: write the premises down first, then let others check whether the conclusion follows from them.

---

3. The System: Five Components

What the five components have in common is that each one shortens the distance between "saying" and "bearing".

This section reconstructs only; it does not evaluate. All text in quotation marks is taken verbatim from the Glacier website.

3.1 The North Slope: Choice of Path

"We take the industry's north slope, seeking the core assets that can lead the next paradigm revolution. The north slope is hard to climb — and it is never congested." (「坚定选择产业的『北坡』，寻找能够引领下一代科技范式革命的核心资产。北坡难爬，但北坡上不堵车。」) [9]

The difficulty of the north slope is defined explicitly as a cost in time, not in intellect: "The difficulty is in time. On the South Slope consensus comes early and so does the price; understanding carries no premium there. The North Slope takes tens of thousands of hours of research and round after round of walking alongside. Money cannot compress that. You can afford the equity. You cannot afford the time." (「难在时间。南坡共识早、价格也贵，认知没有溢价；北坡要几万小时的调研，一轮又一轮的陪跑。这件事没法用钱压缩。买得起股权，买不起时间。」) [10] This is a judgment about the source of scarcity: tradable scarcity (equity) is separated from non-tradable scarcity (time, accumulated understanding), and the firm places itself on the second side.

3.2 The Fact Base: Working Order

"When a company comes to us, the first step is not writing material and not scheduling meetings. It is recovering the facts." (「一家公司来找我们，第一步不是写材料，也不是排会议。是恢复事实。」) "The bottom layer of the Glacier pyramid is called the fact base — facts first, claims second. The fact base is the set of hole cards every judgment shares. The order cannot be reversed." (「庚辛金字塔的底层就叫事实底座——先看事实，再听主张。所谓事实底座，就是所有判断共用的那一组底牌。顺序不能倒。」) [9]

Here there is a discipline concrete enough to be executed:

"Because when something cannot be explained clearly, it is usually not a problem of expression. The facts have not been fully recovered. Whatever can be said can be said clearly — Wittgenstein wrote that in the Tractatus, and we treat it as the prequel to the Glacier philosophy. On the workbench it is hard-edged: the parts you cannot explain are exactly the parts due diligence will check line by line. If you cannot fill it in, do not write it yet." [9] (「因为一件事说不清楚，多半不是表达问题。是事实还没恢复完。凡是可说的，都能够说清楚——维特根斯坦在《逻辑哲学论》里这么写，我们把它当作庚辛哲学体系的前传。但是它在工作台上很硬：说不清的地方，就是尽调（due diligence）里会被逐条核对的地方。填不动，就先别写。」)

This is the only direct citation of a philosophical text in the whole system, and it is the one Section 4 examines most closely.

3.3 Entropy Reduction: A Theory of Value

"Physics has a word for it: entropy. Left alone, everything drifts towards disorder. The financing table is the same. The founder holds the technology piece, the institutions hold the market piece, the existing shareholders hold the history piece. Nobody is lying, and nobody can assemble the whole picture." (「物理里有个词叫熵增，意思是不加干预，一切自发走向散乱。融资桌上也一样。创始人握着技术那一段，机构握着市场那一段，老股东握着历史那一段。没有人说谎，但是谁也拼不出全图。」) [10]

"Reducing entropy means gathering scattered facts into a conclusion you can act on. Deep thinking plus precise execution is the force that injects that order. The force itself is value creation, not a by-product." (「所谓熵减，就是把散落的事实收拢成一个可判断的结论。深度思考加精准执行，就是把秩序注入进去的那道力。其实这道力本身就是价值创造，不是附属品。」) [10]

This passage answers the oldest question in the economics of investment banking: what an intermediary is paid for. Glacier's answer rests not on matchmaking but on reassembly: taking information that is distributed, that no one lies about yet no one holds in full, and reassembling it into a structure a decision can be made on.

3.4 The Camp: Definition of the End

"Arrival is not the end; returning to camp is. Nor is closing the end: only when the company still stands in the next cycle is the journey complete." (「抵达不是终点，回到营地才是。交割也不是终点：公司在下一个周期还站得住，这一趟才算走完。」) [11]

Paired with it is a set of statements about redundancy: "Redundancy first, returns second. Safety ranks far above the rate of return." (「先算冗余，再算收益。安全先于回报。」) "Fair weather or foul, thirty kilometres a day … a sustainable tempo carries the team to the finish; a sprint cannot. Thirty kilometres a day is not fast; the advantage is that there is still a next day." (「不管天气好坏，每天三十公里……可持续的节奏把队伍带到终点，冲刺不能。每天三十公里不算快，好处是第二天还有。」) [11] This discipline of redundancy corresponds to the Antarctic reconstruction (see Figure 1).

And the capacity limit derived from it: "One more mandate is margin withdrawn. Less is more. The number of mandates is strictly limited — to take one more is to withdraw margin from the companies already in hand." (「多接一单，就是抽走余量。少就是多。接案数量严格受限：多接一单，就是从在手的公司身上抽走余量。」) [11] "Judgment, research, coordination and senior attention are non-renewable resources." (「判断、调研、协调与合伙人的注意力，都是不可再生资源。」) [9]

One common internal paraphrase needs correcting. The website says "We do twenty to thirty mandates a year, half of them repeat mandates, existing clients coming back" (「我们每年做二三十个案子，其中一半是老客户复购」) [10], not "only a dozen or so a year" (「一年只接十几个」). This paper follows the website's wording.

3.5 Six Dimensions and Shared Stakes: The Structure of Collaboration

"The model covers six dimensions, trust first: trust, information, narrative, tempo, relationships, accountability." (「该模式涵盖六大核心维度（信任为先）：信任、信息、叙事、节奏、关系、担当。」) [9]

The order is explained explicitly as dependence, not layout: "The other five all grow on it … Which is to say, the first thing delivered under a mandate is not material, it is the hole card. Without trust, what comes out is only the dressed-up version. The softer the foundation, the more dangerous each storey built on it." (「另外五级都长在它上面……也就是说，托管的第一笔交付其实不是材料，是底牌。没有信任，拿出来的只会是修饰过的版本。地基一软，越往上盖越危险。」) [9] This collaborative structure sits on one of the website's four levels (see Figure 2).

Paired with this is co-investment with the firm's own capital: "Only by putting one's own money in (skin in the game) does one truly understand what keeps investors awake; only by working in-house does one stand in the same trench as the founder." (「唯有以自有资金躬身入局（Skin in the game），才能深刻体会投资人的风险与决策逻辑；唯有深度深入现场（In-house），才能与创业者并肩作战在同一个战壕。」) [9]

Structurally these two do the same thing: they turn the adviser from "the one who speaks" into "the one who bears the consequences". This has a clear counterpart in agency theory. Jensen and Meckling showed that when the residual claims of agent and principal are separated, agency costs are endogenous to the structure itself rather than a matter of moral standard [7]. Co-investment with one's own capital is a direct intervention in that structure: it lowers agency costs not by promise but by changing the payoff function.

---

4. The Argument: What the Tractatus Actually Has to Do with This Method

Of four correspondences only one really holds: "whatever can be said can be said clearly" became "if you cannot fill it in, do not write it yet"; the other three are useful metaphors, not arguments.

This section is the core of the paper. Its position is: of four candidate correspondences, only one holds; the other three are analogies. Each is examined in turn. For the text of the Tractatus and its place in Wittgenstein's thought, this paper follows the standard survey literature [4].

4.1 Strong Correspondence: Proposition 4.116

Tractatus 4.116: "Everything that can be thought at all can be thought clearly. Everything that can be said can be said clearly." (「凡是能够被思考的，都能够被清楚地思考。凡是能够说的，都能够说清楚。」)¹[1][2]

Glacier's transposition reads: "The parts you cannot explain are exactly the parts due diligence will check line by line. If you cannot fill it in, do not write it yet." (「说不清的地方，就是尽调里会被逐条核对的地方。填不动，就先别写。」) [9]

This correspondence holds for three reasons.

First, it is a rule of production, not a quotation. In the original, 4.116 is a methodological claim: Wittgenstein uses it to mark out the legitimate working range of philosophy, and its radical version appears in 6.53, where the right method of philosophy is "to say nothing except what can be said, i.e. the propositions of natural science" (「只说可说的东西，即自然科学的命题」) [1]. Glacier's transposition keeps this methodological form: it likewise treats "cannot be said clearly" as a signal to stop the action, not as an occasion for stronger rhetoric. This is the crux. Most business writing, on reaching something it cannot say clearly, adds adjectives; this rule requires the pen to stop.

Second, it can be broken, and is therefore a real rule. The test of whether a discipline really exists is whether it has rejected something that would otherwise have passed. Any sentence that cannot be traced back to the fact base must be deleted, however moving it is.

Third, its failure mode matches that of the original proposition. The difficulty of 4.116 is that it is itself hard to say clearly; this is exactly the self-referential predicament of the "ladder" in 6.54 [1]. Glacier's version has an isomorphic predicament: "cannot be said clearly" is itself a judgment, and who makes that judgment, and by what standard, cannot be fully stated inside the system. The next section returns to this point.

4.2 Weak Correspondence (Analogy): The Picture Theory and the "Fact Base"

Tractatus 2.1–2.225 holds that a proposition is a picture of reality; "In order to discover whether the picture is true or false we must compare it with reality" (「为了发现图像是真是假，我们必须拿它与实在比较」) (2.223); "It cannot be discovered from the picture alone whether it is true or false" (「仅凭图像本身无法发现它是真是假」) (2.224) [1][3].

On the surface this is almost the same sentence as "Recover the facts first, then hear the claims" (「先恢复事实，再听主张」): the truth value of a claim (the picture) can only be settled by comparing it with the facts (reality), not by the claim's own internal coherence.

But this paper holds that this is an analogy, not a lineage, for two reasons.

First, the two "facts" are not the same thing. The state of affairs (Sachverhalt) of the Tractatus is a technical concept within the framework of logical atomism: a configuration of simple objects according to logical form, whose existence is the conclusion of an a priori argument, not the product of empirical inquiry [3]. Glacier's "fact base", by contrast, is a list of empirical material: patents, product matrix, customer orders, team structure, competitive landscape, latent risks and the history of past financing [9]. To call the latter an application of the former is to misread a metaphysical category as a due-diligence checklist.

Second, the feasibility of "comparison" is entirely different. In the Tractatus the comparison of picture with reality is guaranteed possible by logic; in the primary market that comparison is precisely what is scarce. Most key facts about an unlisted company cannot be independently verified by a third party, which is exactly the quality-uncertainty problem described by Akerlof [5] and the signalling problem treated by Spence [6]. What Glacier does (deep in-house presence, long-term walking alongside, co-investment with its own capital) is, in economic terms, closer to a Spence-style costly signal than to a picture-theory truth comparison.

In other words, the theory actually at work here is not Wittgenstein but information economics. The picture theory supplies a memorable metaphor, not an argument. Admitting this is more honest than forcing the two together.

4.3 Weak Correspondence (Analogy): Proposition 7 and "Not Saying"

Tractatus 7: "Whereof one cannot speak, thereof one must be silent." (「凡是不可说的，必须保持沉默。」)²[1][2]

The Glacier system does contain several disciplines of "not saying": "write only what has already happened, never what has not" (「只写已经发生的事，不写还没发生的事」) [10]; and, in the portfolio disclosure, "some companies are omitted under confidentiality agreements or at their own preference" (「个别公司因保密约定或其自身节奏，暂不列出」) [11].

But the two silences are entirely different in kind.

Wittgenstein's silence is a logically necessary silence. What cannot be said (ethics, value, the meaning of the world) is not forbidden from discussion; it simply cannot be meaningfully expressed at all. 6.421 states plainly that "ethics cannot be expressed. Ethics is transcendental" (「伦理是不可表达的，伦理是超越的」) [1]. This silence admits no exception and needs no discipline to maintain it, because breaking it produces not a false statement but meaningless syllables.

Glacier's silence is a silence of prudence and contract: it can be said, but the firm chooses not to say it, because saying it would harm the other party's interests or go beyond known facts. This silence can perfectly well be broken, which is why it has to be written down as a discipline.

To describe the latter as a practice of the former is a category confusion. If a counterpart for this discipline must be found in the history of philosophy, the closer fit is not proposition 7 but the everyday version of the epistemic duty that "a claim must have grounds".

It should be said that this confusion is not made by Glacier's own text; the website never claims to be practising proposition 7. This paper raises it to keep the slippage from being written into future chains of citation. It should also be noted that Wittgenstein himself revised several early positions of the Tractatus in his later work; this paper does not evaluate the system on that basis [13].

4.4 A Correspondence the Website Does Not Mention but Which Structurally Exists: The Ladder of 6.54

"My propositions are elucidatory in this way: he who understands me finally recognizes them as senseless, when he has climbed out through them, on them, over them. (He must so to speak throw away the ladder, after he has climbed up on it.)" (「我的命题以如下方式起阐明作用：理解我的人，当他借助这些命题、循着它们爬出去之后，最终会认识到它们是无意义的。（可以说，他必须在爬上梯子之后把梯子扔掉。）」) [1]

This paper regards this as the proposition most worth Glacier's attention, because it describes a method's self-cancellation.

Any system that writes "recover the facts first" as its first step will in the end meet this moment: once the facts have been recovered completely enough, the decision is still not derived from the facts. Glacier's own text in fact touches this boundary: "Five workstreams, three words: the math, the curve, the people" (「五件事归结起来是三个词：算账、看势、落人」), where "the people" means to "judge the people — decide whether this team can turn a curve on paper into numbers on a financial statement" (「落人，最终判断团队能不能把纸上的曲线，做成报表上的数字」) [9]. "The people" is not calculated. It is a judgment the fact base cannot fully carry.

At this point the system has used up its ladder. Admitting so is closer to Wittgenstein than claiming that facts can be pushed all the way to a conclusion.

---

5. When This Philosophy Does Not Hold, and What It Costs

The dearest cost of this system is that its costs are visible every day, while its benefits live only in the counterfactual.

A paper that does not write this section has no right to call itself a paper. This section lists four costs and four failure conditions.

5.1 Cost One: The Opportunity Cost Priced at "One Deal Fewer" Is Openly Marked

Glacier itself has written this account into its public text:

"We stand back on purpose, to keep the one vote that says 'stop'. Have we used it? We have. What it buys is the nerve to talk the risk through. What it costs is one deal." [10] (「我们刻意站远，是为了保留说『停』的那一票。按过吗？按过。换来的是敢把风险讲透。代价是少一单。」)

This self-statement matters because it restates a choice that could have been called a virtue (prudence) as a measurable loss.

But the account has another side: "one deal fewer" (「少一单」) is observable; "what one more deal would have cost" (「多接一单会损失什么」) is not. The former enters every internal review; the latter never does. When a discipline of this kind gives way inside an organisation, it is usually not because someone opposed it, but because no one could put the losses it avoided on the table.

5.2 Cost Two: Systematic Under-weighting of "Knowledge That Cannot Be Said"

This is the cost this paper regards as the heaviest, and it is a direct side effect of the discipline "whatever can be said can be said clearly".

The Polanyi-style insight that people know more than they can tell corresponds in epistemology to "knowing how", a class of knowledge that resists being put into propositions [12]; in organisational knowledge research, Nonaka developed it into a model of conversion between tacit and explicit knowledge [8]. In early-stage technology companies, the information with the most predictive power sits largely on the tacit side: an engineer's feel for process yield, how fast a team reacts in a post-mortem of a failure, the founder's first sentence on hearing bad news. In the current period none of these can be "filled in" (「填不动」).

A method that prescribes "if you cannot fill it in, do not write it yet" will defer adopting such information until it can be said. Deferral is not discarding, but at an early stage deferral often amounts to missing.

Here is a real tension: the heaviest element in the system, "the people" (Section 4.4), sits entirely on the tacit side, while the system's first discipline demands that things be made explicit. This paper cannot tell from public material how the firm handles this in practice, and so offers no evaluation; it only marks the location.

5.3 Cost Three: The Ceiling on Scale Is Endogenous, Not a Choice

"Judgment, research, coordination and senior attention are non-renewable resources" (「判断、调研、协调与合伙人的注意力，都是不可再生资源」) [9], together with "the partner sits from the first meeting to the day of closing, with no hand-off in between" (「合伙人从第一次会议坐到交割那天，中途不换人」) [11], means that the ceiling on capacity is locked to the number of partners, and that this ceiling cannot be scaled linearly by hiring, because what is delivered is precisely those few people's judgment.

This is not a problem process optimisation can solve; it is the system's defining constraint. Any expectation that it expand significantly with the method unchanged is a demand that it give up one of "deep in-house presence" or "no hand-off in between".

5.4 Cost Four: Being "Refutable" Has a Social Cost of Its Own

"To keep the one vote that says 'stop'" (「保留说『停』的那一票」) is clear as a discipline and expensive as a relationship. Glacier places the foundation of the six dimensions on trust [9], and saying "stop" is precisely a withdrawal from trust. The system writes two things that pull against each other into the same document: trust must be accumulated over the long term, while a veto must be exercised at once. That the two do not conflict is the ideal case; how to choose when they do conflict is not laid down anywhere in the public text. This is a gap in formulation that this paper found, not a contradiction, but it should not be treated as settled either.

5.5 Four Failure Conditions

The four items above are costs, paid even when the system runs normally. The four below are failure conditions: when they occur, the system's premises no longer hold.

First, when the moat is not in the labs and on the production lines. The first axiom is a falsifiable empirical proposition. If the main barrier of a given generation of technology migrates to distribution, data feedback or standard-setting, the "north slope" position remains executable, but what it points to is no longer the core asset. The system stays coherent; it is simply measuring the wrong mountain.

Second, when what the market prices is not facts but the speed at which consensus forms. The premise of "injecting order is value creation" (「秩序注入即价值创造」) [10] is that clearing up the facts can change the price. In a range dominated by capital liquidity and narrative speed, the time needed to clear up the facts may be longer than the window itself. Then "only slowing down makes the structure visible" (「慢下来才看得见结构」) [10] is right, but by the time the structure is visible the price is gone.

Third, when silence is misread as the absence of facts. The restraint formed jointly by "write only what has already happened" (「只写已经发生的事」) [10] and by confidentiality obligations will, in an environment that uses volume as a proxy for capability, be read as a lack of action. The cost of this item is borne not by the firm but by the companies it serves.

Fourth, when "cannot be said clearly" is treated as a veto rather than a to-do. This is the deformation the discipline is most prone to internally. "If you cannot fill it in, do not write it yet" (「填不动，就先别写」) is a rule about the order of writing; once it is simplified in execution into "if you cannot fill it in, do not do it", it degrades from a discipline of truth-seeking into a device for risk avoidance, and the latter will filter out, precisely, every early opportunity that has not yet been put into words — the very ones the north slope is supposed to hold.

---

6. Conclusion

To judge whether a system is philosophical, look not at whom it cites but at whether it dares to write down when it does not hold.

Back to the question of the introduction: is naming after the years in which the Tractatus was written a rhetorical borrowing, or a commitment?

This paper's answer has three layers.

In the sense of intellectual lineage, it is a borrowing, and the firm itself admits as much. "That is the whole provenance" (「来历就这么多」) [10] confines the naming to a coincidence of calendar years and claims no inheritance. If later writers describe Glacier as "the business practice of Wittgenstein's philosophy" (「维特根斯坦哲学的商业实践」), that is the writer's addition, not the original's claim.

In the sense of method, there is one real commitment. Proposition 4.116 has been transposed into a working rule that rejects things; the rule has a cost, can be broken, and its failure mode is isomorphic with the original proposition's (Section 4.1). One real correspondence is more persuasive than four forced ones.

In the sense that matters more: whether a system is philosophical has nothing to do with whom it cites. The criterion is whether it can write its premises in a form that can be overturned, and say under what conditions it does not hold. By that criterion, the most philosophical passage in the Glacier system is not the one that cites Wittgenstein but the one that says "what it costs is one deal" (「代价是少一单」): it marks the price of its own discipline in public.

This agrees with Wittgenstein's real legacy. The Tractatus matters not for what it built but for announcing, in 6.54, that its own propositions would in the end be thrown away. A method willing to describe when it fails inherits that posture, not those sentences.

This is probably also the most honest position a boutique investment bank's philosophical system can reach: it does not claim to be right; it claims only to be checkable.

---

Eleven papers are not eleven topics. They are eleven links on one chain: segment first, then sequence, then take apart the one move in each link that most often goes wrong.

Read nextGI-WP-2026-P11　Borrowed Theories: How Frontier-Technology Concepts Explain the Daily Work of an Investment Bank一篇说自己什么时候不成立，一篇说自己的概念从哪借来。互为脚注。

FOR INVESTORS

For investors递到你面前的，是我们自己敢投的。进去看 →

转人工 · ASK A HUMAN把这个问题交给赵羚茜（Gogo Zhao，投资人关系负责人）安排——她会把它交到对的人手里。 联系庚辛

庚辛人生（Certainty of Money） · 庚辛研究院通讯论文 · GitHub · PDF 全文 · DOI · 可引用，CC BY 4.0。

References

Format: Author (year). Title. Journal/publisher, volume(issue), pages. DOI or accessible link. Philosophical primary texts are cited by edition; the proposition numbers cited in the text were checked one by one against [1].

[1] Wittgenstein, L.（1922）. Tractatus Logico-Philosophicus（C. K. Ogden, trans.）. London: Kegan Paul, Trench, Trubner & Co. Full electronic text (numbered propositions): The Wittgenstein Project, https://www.wittgensteinproject.org/w/index.php/Tractatus_Logico-Philosophicus_(English)（accessed 2026-09-20）. Propositions 2.1, 2.223, 2.224, 4.116, 5.6, 6.53, 6.54, 6.421 and 7 cited in this paper were all checked against this edition.

[2] Wittgenstein, L.（1922）. Tractatus Logico-Philosophicus（electronic edition of the C. K. Ogden English translation）. Project Gutenberg, eBook #5740. https://www.gutenberg.org/ebooks/5740（accessed 2026-09-20）

[3] Klement, K. C.（ed.）. Tractatus Logico-Philosophicus: Side-by-Side-by-Side Edition（German original with the Ogden and Pears–McGuinness English translations in parallel）. University of Massachusetts Amherst. https://people.umass.edu/klement/tlp/（accessed 2026-09-20）

[4] Biletzki, A., & Matar, A.（2002, revised 2026）. Ludwig Wittgenstein. The Stanford Encyclopedia of Philosophy. Stanford University. https://plato.stanford.edu/entries/wittgenstein/（accessed 2026-09-20）

[5] Akerlof, G. A.（1970）. The Market for "Lemons": Quality Uncertainty and the Market Mechanism. The Quarterly Journal of Economics, 84(3), 488–500. https://doi.org/10.2307/1879431

[6] Spence, M.（1973）. Job Market Signaling. The Quarterly Journal of Economics, 87(3), 355–374. https://doi.org/10.2307/1882010

[7] Jensen, M. C., & Meckling, W. H.（1976）. Theory of the Firm: Managerial Behavior, Agency Costs and Ownership Structure. Journal of Financial Economics, 3(4), 305–360. https://doi.org/10.1016/0304-405X(76)90026-X

[8] Nonaka, I.（1994）. A Dynamic Theory of Organizational Knowledge Creation. Organization Science, 5(1), 14–37. https://doi.org/10.1287/orsc.5.1.14

[9] Glacier Institute (2026). 庚辛研究院｜Glacier Institute [Glacier Institute] [in Chinese]. Glacier Capital official website. https://glacier.mba/institute.html (accessed 2026-09-20). The passages cited in this paper on "founding intent and the two axioms", "fact base", "six dimensions of full mandate", "Skin in the game / In-house" and "limited service capacity" are all taken verbatim from that page.

[10] Glacier Institute (2026). 卷一·哲学 系列（「北坡难爬，但北坡上不堵车」「名字即长跑」「把事实按顺序放好」「秩序不会自己出现」）[Volume One: Philosophy series (The North Slope Is Harder, but It Is Never Crowded; The Name Is Already a Long Run; Put the Facts in Order; Order Does Not Appear on Its Own)] [in Chinese]. Glacier Capital official website, Glacier Institute page. https://glacier.mba/institute.html (accessed 2026-09-20)

[11] Glacier Capital (2026). 科创时代的确定性（官网首页全文，含「20 / THE AMUNDSEN DISCIPLINE」「接案的门槛」「走得回来才算赢」各章）[Certainty in the Sci-Tech Era (full text of the website home page, including the chapters "20 / THE AMUNDSEN DISCIPLINE", "The Bar for Mandates" and "Winning means making it back")] [in Chinese]. Glacier Capital official website. https://glacier.mba/index.html (accessed 2026-09-20)

[12] Pavese, C.（2021）. Knowledge How. The Stanford Encyclopedia of Philosophy. Stanford University. https://plato.stanford.edu/entries/knowledge-how/（accessed 2026-09-20）. Background reference for Section 5.2 on "knowing more than one can tell".

[13] Candlish, S., & Wrisley, G.（1996, revised 2019）. Private Language. The Stanford Encyclopedia of Philosophy. Stanford University. https://plato.stanford.edu/entries/private-language/（accessed 2026-09-20）. Background reference for Section 4.3 on the later Wittgenstein's revision of his early positions.

On the source of the Amundsen–Scott narrative in [11]: the Glacier website states explicitly that the story "comes from a 2012 talk by Wang Xing; the chart is a schematic reconstruction, not a measured track" (「均出自王兴 2012 年的一次演讲；图为示意推演，不是实测航迹」). This paper treats it as a second-hand citation and does not use it as primary evidence in history or polar exploration.

---

Glacier Institute 庚辛研究院. North Slope and Base Camp: a boutique bank’s philosophical system and its limits of language[R]. Glacier Institute Working Paper GI-WP-2026-P6-EN. 2026. DOI:10.5281/zenodo.22869201.

Glacier Institute. (2026). North Slope and Base Camp: a boutique bank’s philosophical system and its limits of language (Glacier Institute Working Paper No. GI-WP-2026-P6-EN). Zenodo. https://doi.org/10.5281/zenodo.22869201

@techreport{glacier2026p6en,
  author      = {{Glacier Institute}},
  title       = {{North Slope and Base Camp: a boutique bank’s philosophical system and its limits of language}},
  institution = {Glacier Institute},
  type        = {Working Paper},
  number      = {GI-WP-2026-P6-EN},
  year        = {2026},
  doi         = {10.5281/zenodo.22869201},
  url         = {https://doi.org/10.5281/zenodo.22869201}
}