Keeping the Direction
守向
The earlier papers wrote what a company oriented toward the Kingdom of Ends does toward five parties. This paper asks what keeps that direction. In February 2020, Wells Fargo paid 3 billion dollars, through a deferred prosecution agreement and a civil settlement, to resolve the criminal and civil investigations into its sales practices; it admitted that from 2002 to 2016, under the Community Bank's onerous sales goals and accompanying management pressure, customers were charged fees and interest they should not have been charged. Internal investigators had called it a "growing plague" as early as 2004, yet the Community Bank's senior leaders did not change that sales model. The problem had been seen early; why was it not changed? This paper does not answer for this case; what it asks is what keeps a company's direction. Consensus loosens: people change, companies are acquired, and a written consensus slowly comes to exist only on paper; and a company cannot require people to agree with it. So keeping the direction cannot presuppose consensus, and can rest only on constructs and positions: the baseline is written into the articles and into decision procedures, and changing it must be public and open to questioning; self-correcting channels are independent of the party questioned and are examined by the three questions, and those who ask are not penalized for asking; incentive structures do not override the baseline, and the position that sets a target is accountable for where the target presses (GS12). The John Lewis Partnership Council's power to dismiss the chairman, the Delaware public benefit corporation's periodic reports and the stockholders' power to sue, and the external channels run by competent authorities and the ban on retaliation in the EU Whistleblower Directive each show one kind of arrangement; John Lewis's plan to shrink its Partnership Council, and the third generation of Crossways, in this series' constructed examples, changing the composition of the Seller Appeals Committee, show that the channels that keep the direction can themselves be changed. The judgment of hard times has reasons that can be given and can be questioned; if the judgment is wrong, the position that made it bears the corrections and the consequences of accountability that correspond to it, and individual responsibility beyond that requires independent grounds (GS13). Finally the paper speaks of direction: keeping the direction is also leaving conditions for those who come later. The paper has two propositions (GS12, GS13), each stated at the end of its section and listed together at the end.
Writing Declaration: This paper was co-drafted with Claude (Anthropic). All intellectual decisions, framework design, and final editorial judgments were made by the author.
Introduction
On February 21, 2020, the U.S. Department of Justice announced that Wells Fargo had agreed to pay 3 billion dollars to resolve the criminal and civil investigations into its sales practices. In law this was a deferred prosecution agreement and a civil settlement, not a criminal conviction. Among the facts the company admitted: from 2002 to 2016, under "The Community Bank's onerous sales goals and accompanying management pressure," the company "collected millions of dollars in fees and interest to which the Company was not entitled," harmed the credit ratings of some customers, and unlawfully used customers' sensitive personal information (U.S. Department of Justice press release, 2020-02-21).
The problem had been seen early. According to the same material, internal investigators had called these practices a "growing plague" as early as 2004, and in 2005 said they were "spiraling out of control"; yet the Community Bank's senior leaders "refused to alter the sales model."
The problem had been seen early; why was it not changed? This paper does not answer for this case. What it asks, from here, is what keeps a company's direction.
What is judged here is this practice and its structure, not the whole bank, and minds are not read. §1 says that consensus loosens; §2 takes up the arrangements that keep the direction; §3, judging hard times and answering for the judgment; §4 speaks of direction.
The reading follows the Prequel: the construct takes the ought, the position takes accountability, the person takes his own law.
§1 Consensus Loosens
The Prequel said that consensus is identified by where it lands: written into the articles and into decision procedures, able to be invoked when decisions are made, with positions holding their accountability by it. Even once it has landed, it can loosen. People change, CEOs change, companies are acquired, hard times come. A written consensus slowly comes to exist only on paper: the sentence is still in the articles, but no one invokes it when deciding; or it is invoked, but no one uses it to question anything.
There is one more thing that makes keeping the direction harder. A company cannot require employees to agree inwardly with the consensus written into the company: "A construct may require a position to do something; it may not require a person to think something." (F6 §3.3; Paper 2, GS6) This proposition protects people, and this series does not take it back. But it has a consequence: keeping the direction cannot presuppose that "everyone still agrees." New employees may not agree, the next CEO may not agree; if the direction is still to be kept, it can rest only on constructs and positions (own undertaking).
One more thing must be kept apart. A company that has deliberately changed its direction is outside this series' scope (Paper 1, §3). What this paper speaks of is the kind that no one meant to change, but that has slowly loosened.
§2 The Arrangements That Keep the Direction
Normative layer. Of the arrangements that keep the direction, this series requires only four things.
First, the baseline is written into the articles and into decision procedures. What is written in, if changed, must be changed publicly, and can be questioned.
Second, self-correcting channels are independent of the party questioned. Independence is examined by Institution Theory's three conditions: the channel "is independent of the party questioned, internal subjects can actually initiate it, and initiating it brings no retaliation" (Institution Theory Z34). Whether the channel has efficacy is examined by the three questions: who can activate it; what actual efficacy it has on reasons, actions, rules, or the holding of office; how the party questioned cannot cancel it on its own (Political Economy Paper 3, E16). Where the party questioned can close on its own the route that questions it, this is read into the second gate (Political Economy Paper 4, E22).
Third, those who ask are not penalized for asking (Institution Theory Z34).
Fourth, incentive structures do not override the baseline. That a company sets targets for a position is not in itself the problem; where those targets press is the problem. Political Economy writes that the rate of return set in the place of the sole goal "provides a signal to be examined; if it further acquires a standing that overrides guarding the line, cancels the questioning and change of direction of those concerned, or decides for the investee its own ends, it is judged by these actual effects" (Political Economy Paper 5, E27). The original speaks of the return targets an investor sets for the construct invested in; here it is changed to the targets a company sets for a position (own undertaking). The third of the original's three branches, "decides for the investee its own ends," cannot be extended to a position: a position has accountability, not ends of its own; here it is read as deciding for the person behind the position his own affairs (Paper 2, GS6). That the position that sets a target is accountable for where the target presses is not in E27's original sentence; its ground lies in the accountability of positions (Political Economy Paper 1, E2; Institution Theory Z6; own undertaking).
Contrast: being asked does not prove the channel effective. In the Wells Fargo case, internal investigations had already pointed out the problem. The Department of Justice's material says: "Despite knowledge of the illegal sales practices, Community Bank senior leadership failed to take sufficient action."
What this case shows is: being asked does not prove the channel effective. Who could compel correction, and whether the party questioned could cancel the necessary effects, require a further examination of structure. The first of the three questions has an answer here: internal investigators could activate it, and did. What remains to be examined are the latter two: what actual efficacy asking had, and whether the party questioned could stop it on its own. The material this paper has examined reads only as far as "being asked does not prove the channel effective," and gives no answer to these two questions. The Department of Justice's material writes the Community Bank's senior leaders, the group's management and the board separately; nor are they pressed here into one undifferentiated "party questioned": the latter two questions must be examined layer by layer.
Examples at the rule layer.
John Lewis Partnership has a Partnership Council elected by employees. Under Rule 10 of its current constitution (the August 2026 edition), "If the Council judges that the Chairman has failed to fulfil, or is no longer a suitable person to fulfil, the responsibilities of the office as laid down in the Constitution, it may pass a 'Resolution upon the Constitution' to dismiss the Chairman according to the Articles of Association of John Lewis Partnership Trust Limited." In May 2023 the Council held a confidence vote on the then chairman's leadership; that vote was not itself binding (ITV News, 2023-05-10). Dismissal is a channel in the constitution; the confidence vote is something else; the two are read separately.
A Delaware public benefit corporation "shall no less than biennially provide its stockholders with a statement" setting out the public benefit objectives the board has set, the standards adopted to measure progress, objective factual information based on those standards, and an assessment of how far the objectives have been achieved (Delaware Code, Title 8, §366(b)); stockholders owning together at least two percent of the corporation's outstanding shares (for a listed corporation, two percent or shares with a market value of two million dollars, whichever is less) can sue to enforce the balancing requirement (§367). Read by the three questions: who can activate it, stockholders who meet the threshold can; on the basis of being an employee, a user or a member of society alone, one cannot activate this suit (own undertaking). These standings may overlap with being a stockholder who meets the threshold; nor does this provision sum up other channels of questioning. Whether the statement is made public may be provided for separately by the certificate of incorporation or the bylaws.
The EU's Whistleblower Protection Directive requires that "legal entities in the private sector with 50 or more workers" set up internal reporting channels (Art. 8(3)); that a reporting person may report first through internal channels, "or by directly reporting through external reporting channels" (Art. 10); that reports made through external channels are received and followed up by authorities designated by Member States ("Member States shall designate the authorities competent to receive, give feedback and follow up on reports") (Art. 11(1)), with Article 12(1) further defining the "independent and autonomous" character of external channels by standards such as the completeness, integrity and confidentiality of information; and it prohibits "any form of retaliation," suspension and dismissal included (Art. 19) (Directive (EU) 2019/1937). This case shows two things: external channels are run by competent authorities, not by the entity reported on, and so are independent of the party questioned; and those who ask are not penalized for asking.
Two things used earlier are also arrangements that keep the direction: Patagonia placing its voting shares in a trust set up to safeguard the company's purpose, and Bosch separating voting rights from earnings (Paper 1).
Judgment layer: the channels that keep the direction can themselves be changed. A June 2025 report (FashionUnited, citing The Daily Telegraph) said that John Lewis was preparing to cut its Partnership Council from fifty-seven members to forty-three; the reason the company gave employees was that the existing structure relied "too heavily on hierarchy and escalation." This is a plan in a report; whether it has taken effect this paper has not examined.
This paper does not judge this adjustment. Whether a council should have fifty-seven members or forty-three is not something the normative layer can answer. Only three things are asked here: whether the decision to change the channel is public; whether it can be questioned; whether, among those asking, there is anyone other than the party questioned. The channels that keep the direction are constructs too, and must also be open to questioning.
Among the examples set up in this series there is also a case that fits "people change." At Crossways Group in Apple City, in the summer of Year 24, the second generation of the Qiao family retired, the third generation took over as chair of the Family Office and chair of the Crossways board, and the voting rights of the Class B shares were thereupon exercised by him. In Year 25 the board chaired by the third generation stopped the Small Shop Starter Program, which had run since Year 15; and the two members of the Seller Appeals Committee elected by and from among the Gold Sellers were changed to appointment by management. What the previous generation had loosened was taken back in the hands of the next; and the one that changed the composition of the channel was the party questioned itself. These two are not one kind of thing: the Small Shop Starter Program is giving conditions, and belongs to direction; the two seats on the appeals committee are a channel of questioning, of the kind that the second requirement in this section requires to be independent of the party questioned. This case shows one thing: if a channel of questioning can be changed at will by the party questioned, then when people change, it can be taken back (Political Economy Paper 3, E16). It is not the slow loosening that §1 speaks of; the party questioned changed the channel directly; nor does it show what kind of person the third generation is, or that every route of questioning at Crossways has been closed (Paper 4, §4).
GS12 Consensus loosens: the arrangements that keep the direction do not presuppose consensus. A company cannot require people to agree, so keeping the direction rests only on constructs and positions: the baseline is written into the articles and into decision procedures, and changing it must be public and open to questioning. Self-correcting channels are independent of the party questioned and are examined by the three questions; those who ask are not penalized for asking. Incentive structures do not override the baseline: the position that sets a target is accountable for where the target presses (per Institution Theory Z34, Political Economy E16, E22, E27, How Is Institution Possible §3.3; own undertaking).
§3 Judging Hard Times, and Answering for the Judgment
Paper 2 said that the cost of hard times falls on what is done; but whether the hard times are real, and how the cost is divided, who has the say?
Normative layer. Whether the hard times are real, how much is borne and for how long: the reasons can be given, and can be questioned. Who decides and who is accountable is written down clearly; the CEO is among them.
How a wrong judgment is borne. Judgments of hard times can be wrong: what was thought impassable could in fact be passed; it was thought that one round of layoffs would be enough, and then another was needed. If the judgment is wrong, the position that made it explains its reasons, submits to review, and bears the corrections and the consequences of accountability that correspond to the position. The construct's account still belongs to the construct. Whether an individual bears responsibility beyond that requires independent grounds, and is not automatically transferred onto some person because the outcome shows the judgment wrong; the CEO is examined in the same way. Political Economy has written: "the construct's account, the position's accountability and individual responsibility with independent grounds are examined separately: the construct's name does not shield the last, and the construct's failure is not automatically transferred onto any one person" (Political Economy Paper 1, E2); Institution Theory writes: "The construct's name does not shield an individual responsibility that has independent grounds, and the construct's failure is not automatically transferred onto any one person" (Institution Theory Z6). The two sentences are two sides of one line: the position bears what is the position's, and the person bears the share that has independent grounds.
Tendencies at the rule layer. What to do in hard times is written down in advance. Lincoln Electric guaranteed employment from 1958, and employees agreed to accept necessary adjustments to posts and schedules (Paper 2, §3): how hard times would be borne was agreed before they came. The determination that hard times exist can be reviewed by a party independent of management (tendency; own undertaking). The two tendencies of Paper 2, §3 also apply here: sharing should be progressive, with the CEO among those who share; and employees take part in deciding how to share.
Example. Writing about that hard time, Barry-Wehmiller's CEO said that at first "I was poised to react as I always had," and then changed to company-wide unpaid furloughs (the Prequel, Introduction). A judgment can change; and he stated the reason for the change.
Contrast. In Year 25 Bluekiln Porcelain Works shut two kiln lines and laid off nine hundred people; the company said cheap porcelain from elsewhere had taken its market. A reason was given. Whether it could be questioned, and by whom, the case does not record. This is exactly what this section asks: once a reason has been given, is there someone who is not the party that judged who can use it to ask, and does asking have any effect?
Judgment layer. Who determines that hard times exist; who reviews afterwards; if the judgment is wrong, what the position bears, and whether an individual has independent grounds for bearing further responsibility.
GS13 Judging hard times, and answering for the judgment. Whether hard times are real, how much is borne and for how long, who decides and who is accountable: these are written down clearly, and the CEO is among them. The reasons can be given and can be questioned; if the judgment is wrong, the position that made it bears the corrections and the consequences of accountability that correspond to it; individual responsibility beyond that requires independent grounds and is not automatically transferred because the judgment was wrong (per Political Economy E2; Institution Theory Z4, Z6).
§4 Direction: Keeping the Direction Is Also Giving Conditions
Keeping the direction is leaving conditions for those who come later: new employees, the next CEO, later users and partners. When they come in, this company is still a company oriented toward the Kingdom of Ends, not because everyone still remembers that sentence, but because that sentence is still in the articles, in the decision procedures, in places where it can be taken up to ask.
Make "open to questioning" the company's own habit: give reasons publicly, review afterwards, keep a record of corrections. A company's record of corrections is the most solid condition it gives to those who come later: they know what has gone wrong here, how it was corrected, and what they can take up to ask with in future.
Two disciplines. The arrangements that keep the direction do not guarantee that it is kept; outcomes are not taken as rules. There will always be times when it is not kept. The Specialist Paper says that the normative layer acknowledges that it has a remainder, and that the remainder yields no permission (A Way Out, a Way Forward, E41); extended to the arrangements that keep the direction (own undertaking): where a set of arrangements does not reach, that it does not reach there does not make anything permissible.
Where This Paper Stops
This paper has laid down two propositions: consensus loosens, and the arrangements that keep the direction do not presuppose consensus (GS12); judging hard times, and answering for the judgment (GS13).
What it has not proved: the particular forms of corporate governance; the design of whistleblowing law; the legal arrangements for acquisition and succession.
With this, the five parties and keeping the direction have been written for the whole series. How it interfaces with existing theories of the company, what it has handed on, and what accounts it still holds itself, are handed to Paper 0.
Remainders left: at what point an incentive structure counts as "overriding the baseline," the normative layer has given a direction, and the judgment must be made case by case by actual effect (§2); when the channels that keep the direction are changed, which ways of changing them themselves cross the line this paper has given only three questions, not a list (§2); who reviews the determination of hard times is written only as a tendency (§3).
The Two Propositions of This Paper
GS12 Consensus loosens: the arrangements that keep the direction do not presuppose consensus. A company cannot require people to agree, so keeping the direction rests only on constructs and positions: the baseline is written into the articles and into decision procedures, and changing it must be public and open to questioning. Self-correcting channels are independent of the party questioned and are examined by the three questions; those who ask are not penalized for asking. Incentive structures do not override the baseline: the position that sets a target is accountable for where the target presses (per Institution Theory Z34, Political Economy E16, E22, E27, How Is Institution Possible §3.3; own undertaking).
GS13 Judging hard times, and answering for the judgment. Whether hard times are real, how much is borne and for how long, who decides and who is accountable: these are written down clearly, and the CEO is among them. The reasons can be given and can be questioned; if the judgment is wrong, the position that made it bears the corrections and the consequences of accountability that correspond to it; individual responsibility beyond that requires independent grounds and is not automatically transferred because the judgment was wrong (per Political Economy E2; Institution Theory Z4, Z6).
References
All SAE works below are cited by concept DOI, which always resolves to the latest version.
I · This system
Qin, H. (2026). How Is Institution Possible: From Inter-Ontological Remainder to Co-Constructive Framework (SAE Foundational Paper 6) [Preprint]. Zenodo. https://doi.org/10.5281/zenodo.19328662
Qin, H. (2026). To Whom the Ought Is Addressed (SAE Institution Theory, Paper 1) [Preprint]. Zenodo. https://doi.org/10.5281/zenodo.23269635
Qin, H. (2026). Open to Questioning (SAE Institution Theory, Paper 6) [Preprint]. Zenodo. https://doi.org/10.5281/zenodo.23271324
Qin, H. (2026). Accumulation (SAE Political Economy, Paper 1) [Preprint]. Zenodo. https://doi.org/10.5281/zenodo.23874138
Qin, H. (2026). Checks (SAE Political Economy, Paper 3) [Preprint]. Zenodo. https://doi.org/10.5281/zenodo.23874142
Qin, H. (2026). Conversion (SAE Political Economy, Paper 4) [Preprint]. Zenodo. https://doi.org/10.5281/zenodo.23874146
Qin, H. (2026). Growth (SAE Political Economy, Paper 5) [Preprint]. Zenodo. https://doi.org/10.5281/zenodo.23874150
Qin, H. (2026). A Way Out, a Way Forward (SAE Political Economy, Specialist Paper) [Preprint]. Zenodo. https://doi.org/10.5281/zenodo.24021385
Qin, H. (2026). What a Company Is Not (SAE Theory of the Company, Prequel) [Preprint]. Zenodo. https://doi.org/10.5281/zenodo.24045716
Qin, H. (2026). Investment (SAE Theory of the Company, Paper 1) [Preprint]. Zenodo. https://doi.org/10.5281/zenodo.24045721
Qin, H. (2026). Working Together (SAE Theory of the Company, Paper 2) [Preprint]. Zenodo. https://doi.org/10.5281/zenodo.24045723
Qin, H. (2026). Partners (SAE Theory of the Company, Paper 4) [Preprint]. Zenodo. https://doi.org/10.5281/zenodo.24045733
II · Statutes and official announcements (originals, tier A)
U.S. Department of Justice. (2020, February 21). Wells Fargo Agrees to Pay $3 Billion to Resolve Criminal and Civil Investigations into Sales Practices [Press release, archived]. https://www.justice.gov/archives/opa/pr/wells-fargo-agrees-pay-3-billion-resolve-criminal-and-civil-investigations-sales-practices
John Lewis Partnership. (2026). Constitution of the John Lewis Partnership (August 2026 edition), Rule 10. https://www.johnlewispartnership.co.uk/~/media/Files/J/john-lewis/corp/documents/Constitution-of-the-John-Lewis-Partnership-Aug-2026.pdf
Delaware Code, Title 8, §§366, 367. https://delcode.delaware.gov/title8/c001/sc15/index.html
Directive (EU) 2019/1937 (Whistleblower Protection Directive), Arts. 8, 10, 11, 12, 19. https://eur-lex.europa.eu/legal-content/EN/TXT/HTML/?uri=CELEX:32019L1937
III · Reports (tier B)
ITV News. (2023, May 10). [Report on the John Lewis chairman facing a confidence vote of the Partnership Council]. https://www.itv.com/news/meridian/2023-05-10/john-lewis-boss-to-face-confidence-vote-over-leadership
FashionUnited. (2025, June 10). [Report on John Lewis's plan to shrink its Partnership Council, citing The Daily Telegraph]. https://fashionunited.uk/news/business/john-lewis-to-trim-staff-committee-amid-turnaround-efforts/2025061082126
IV · Constructed examples
Bluekiln and Bluekiln Porcelain Works, and Crossways Group of Apple City and the Qiao Family Office, are carried over from the examples set up in the Political Economy series; they are not real places or companies.
Interfaces
Main upstream: Institution Theory Z4, Z6, Z34; Political Economy Paper 1, E2, Paper 3, E16, Paper 4, E22, Paper 5, E27; F6 §3.3; A Way Out, a Way Forward, E41.
Examples: the settlement of the Wells Fargo sales practices case (A, facts admitted by the company; a deferred prosecution agreement and a civil settlement); the John Lewis Partnership Council's power of dismissal (A), the confidence vote of 2023 (B) and the plan to shrink the Council reported in 2025 (B); Delaware §§366, 367 (A); the EU Whistleblower Protection Directive (A); Lincoln Electric and Barry-Wehmiller (used in earlier papers, B); Bluekiln Porcelain Works' production cut, and Crossways' third generation stopping the starter program and changing the composition of the appeals committee (constructed).
This series: the Prequel (consensus identified by where it lands); Paper 1 (changing the direction takes a company out of scope; Patagonia, Bosch); Paper 2, GS5, GS6, GS7; Paper 4, §4 (Crossways' Seller Appeals Committee).
Handed to Paper 0: interfacing, handover and remainders.