Honesty, Openness, Trustworthiness, and Secrets

post by NormanPerlmutter · 2023-03-06T09:03:12.139Z · LW · GW · 0 comments

Contents

  Naive Theory of the Virtue of Honesty-and-Openness
  Three Types of Secrets
    Where the Reciprocity Principle Works
    Key Insight -- Where the Reciprocity Principle Fails
  Understanding my errors and biases
  Reassessment of Openness, Honesty, and Trustworthiness
    Naive Venn Diagram -- nested
    Revised Venn Diagram -- all combinations are possible
    Some working definitions
    Examples of each region of the Revised Venn Diagram
  Closing Thoughts
None
No comments

Epistemic status: Uncertain, under development. 

I began working on this essay in December 2021 and then set it aside for a while, made some revisions in December 2022, and some more revisions in March 2023 before publishing it. I feel like I'm still missing some important points and oversimplifying a lot of complex topics. I have tried to allude to some of these complexities in the endnotes.

I welcome feedback from the community on this work in progress.


This essay is about noticing my confusion [? · GW] and seeking to make my beliefs more coherent by carefully adjusting my deeply held values. 

For my whole adult life, I have considered communicating with honesty and openness with my close friends to be among my core values. My System 1 would say that I chose this path because honesty is a virtue and openness goes together with honesty, and being honest and open feels right and good and easy for me. My System 2 would justify the virtue as follows.[1]

Naive Theory of the Virtue of Honesty-and-Openness

1) Consistent world model: Being honest externally helps me to be honest internally and have a clear world-view. In contrast, being externally dishonest is mentally taxing and would require me to keep track of what I told to whom and generate different world models for talking to different people, which can encourage internal dishonesty and a confused world-view.

2) Positive affect: Being honest and open feels good. In my experience, the more I share and the more honestly I share, the better I feel.

3) Signaling trustworthiness: By being honest and open, even/especially when there are some undesirable consequences, I signal strongly that I am an honest and open person, and therefore trustworthy. This has two key benefits which greatly outweigh any undesirable consequences:

  1. Reciprocity Principle: By being open and honest towards others, I encourage others to be open and honest towards me. This spreads the virtue of honesty and openness and also feeds back into benefits (1) and (2) -- helping me to build a consistent world model and spreading positive affect.
  2.  People will believe my word. In particular, if people see that I am open about expressing criticism and disagreement, they will trust me when I express agreement or compliment them. They will know that I really mean it; I'm not just doing it to build them up.

This framework has often, but not always, served me well. In recent years, reflecting on some interpersonal challenges that I've dealt with over the years, I noticed some confusion around this framework, and I sought to explore it more deeply. I had a vague sense that the confusion had something to do with secrets -- because I value honesty and openness, I don't like keeping secrets. In general, though not always, I try to avoid putting myself in situations where I need to keep secrets. But sometimes keeping secrets can be valuable.

Around December 2021, I began thinking seriously about the topic of secrets and my confusion around it. I had some discussions with a few close friends regarding how they handled and thought about secrets. I also did some reading on LessWrong on the topics of secrets and honesty, and especially appreciated Raemon's sequence on Privacy Practices [? · GW]. 

Through these discussions with my friends, I formulated the following theory of three types of secrets and noticed a number of blind spots in my reasoning around honesty/openness/trust. While the ideas below may sound obvious and were things that I was aware of on some level, I had not integrated my ideas about secrets together with my ideas about openness and honesty -- my beliefs and values were inconsistent or at least had unresolved tensions. 

Three Types of Secrets

Consider two friends or allies, "me" and "you." Both people know various secrets that they are considering sharing with each other or hiding from each other. Many of these secrets can be classified roughly into three types, based on who the secret is about and who would be harmed if the secret were shared.[2]

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - 

Type 1 secrets (First-person secrets)
My type 1 secrets are secrets that I know that could harm me if they are shared inappropriately. They are in some sense facts about or pertaining to me.

Type 2 secrets (Second-person secrets)
My type 2 secrets are secrets that I know, but that you do not know, that could harm you if they are shared inappropriately.

These would typically be things like my judgments about you, my feelings towards you, my feelings about certain activities that we do together, etc., which might hurt your feelings or damage our relationship if shared, especially if shared without appropriate tact.

Type 3 secrets (Third-person secrets)
My type 3 secrets are secrets that I know that are also known to a third person, and that could harm that third person if they are shared inappropriately. They are in some sense facts about or pertaining to the third person.

Similarly, you may also have type 1, 2, and 3 secrets that you might keep from me. If shared inappropriately, your type 1 secrets could harm you, your type 2 secrets could harm me, and your type 3 secrets could harm a third party

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - 

Where the Reciprocity Principle Works

Type 2 secrets are in a sense the home court of the Reciprocity Principle of openness and honesty. If my friend and I both want to have an open and honest relationship with each other, we can signal this by sharing our type two secrets with each other. If we are able to do so with appropriate tact and clarity (though this may not be easy), then the relationship will typically improve as a result, and we will accomplish our shared goal of increased transparency.[3]

The Reciprocity Principle also works when considering type 1 secrets by themselves. If I share my type 1 secrets with you, then you might be more comfortable sharing your type 1 secrets with me. 

Key Insight -- Where the Reciprocity Principle Fails

However, another important way to signal safety of sharing type 1 secrets has to do with the relationship between type 1 and type 3 secrets. The signalling for type 1 and type 3 secrets, considered together, works very differently from the Reciprocity Principle.

If you share a type 1 secret with me, then in the context of my interactions with a third party, your type 1 secret becomes my type 3 secret. Therefore, I can signal that I could be a good person for you to share your type 1 secrets with by exercising discretion in how I share my type 3 secrets with you (and with others). If I share my type 3 secrets in a way that you would not want your type 1 secrets to be shared, this could encourage you not to share your type 1 secrets with me. Conversely, if I keep my type 3 secrets private from you or share them only in ways that you consider appropriate, this signals that I could be a safe person to share your type 1 secrets with. 

This type of signaling for type 1 and 3 secrets goes completely against the reciprocity principle. Instead of encouraging you to be transparent with me by being transparent with you, I do the opposite. I can signal that it's safe for you to be transparent with me by keeping secrets from you.[4] Realizing this felt like a major insight to me. 

When I shared this insight with a friend, the friend was quick to point out, "That's obvious. What you're saying is that if you go blabbing people's secrets, then nobody will want to share their secrets with you." Stated that way, it does indeed sound obvious and was a principle that I was familiar with. But it was in my blind spot in connection with my theory of openness and honesty and poked a hole in that theory.

Understanding my errors and biases

Confronted with the analysis above, I sought to adjust my naive theory of openness and honesty. I realized that part of what was going on was that I had glommed together three related, but different virtues: honesty, openness, and trustworthiness. All three are important values to me. All three of them mean different things to different people and in different contexts, and can be further factored down into more sub-virtues.

Part of what I was aiming to do by being open and honest was to signal trustworthiness, and it had fallen in my blind spot that another aspect of trustworthiness is being good with confidentiality (anti-openness) -- being good at keeping secrets and a safe person with whom to share secrets. By sharing a lack of confidence in my ability to keep secrets and by sometimes refusing to keep them when asked, including at times when secrets were shared with me accidentally or shared under an implicit assumption of privacy, I was maximizing openness and honesty, but sometimes at the cost of trustworthiness. 

In some cases [LW · GW], it may even be necessary to lie to protect an important secret -- so not only can un-openness be a signal of trustworthiness, sometimes even dishonesty can be a signal of trustworthiness! Taking things a step further, trustworthiness can be seen as a signal of honesty, so there could even be cases where lying signals honesty (indirectly, by signaling trustworthiness).[5]

Previously, I had been fairly rigid in my honesty value, and proud of that rigidity. I felt that making exceptions to the honesty value would pose a risk to my reputation for honesty (both internally in my self-image and externally in my social/professional network), with unacceptable consequences. However, now I realize that I was perhaps protecting my reputation for honesty, but at potential cost to my reputation for trustworthiness, which I also value highly. 

How did I allow this to happen? I think in part, Automatic Norm Bias prevented me from carefully examining the interactions between these three norms/virtues and the effects of framing them in different ways. 

This Automatic Norm Bias happened in conjunction with a sort of motte-and-bailey fallacy.

Bailey: Openness and honesty
Motte: Honesty

The fallacy was exacerbated by my blind spot around the fact that the openness virtue is not a strict subset of the honesty virtue. (See Naive Venn Diagram in the next section.) So it didn't feel like a motte-and-bailey to me.

And so I not only automatically protected my value of honesty (which I think I am right to value strongly with only very limited exceptions) but also my value of openness (one which I still value but would like to re-assess more thoroughly).

Reassessment of Openness, Honesty, and Trustworthiness

Naive Venn Diagram -- nested

In my naive view, these three virtues were properly nested, as in the Venn Diagram below.

Revised Venn Diagram -- all combinations are possible

In a my revised view, the virtues are not nested. In fact, depending on the context and how the virtues are defined, there are examples to fit every region in the Venn Diagram below. (The diagram is not drawn to scale -- some regions actually have more common examples than others.) I will provide an example for each region, after providing some working definitions for the terms.

Some working definitions

Here are some working definitions of openness, honesty, and trustworthiness. All these terms can be construed narrowly or broadly.

Openness
Not keeping secrets, transparent communication.

More broadly:
Seeking to share as much information as possible.

 

Honesty
Telling the truth, not telling lies, with regard to facts.

More broadly:
Telling the truth not just with regard to facts, but also with regard to things you agree to do -- not making false promises.

More broadly in a different direction: 
Not intentionally deceiving people, not even through truthful statements.

More broadly still:
Not misleading people with your statements, not even unintentionally through truthful statements. Making sure that your ideas are communicated clearly and precisely (or with a level of precision that you specifiy) as well as truthfully.

 

Trustworthiness
Being somebody people can count on to tell the truth

More broadly:
Being somebody people can count on to tell the truth and to keep private information private.

More broadly still:
Being a person of your word -- keeping your commitments and obligations in addition to telling the truth about factual matters.

 

Examples of each region of the Revised Venn Diagram

Not open, not honest, not trustworthy
Telling a lie to conceal information in order to commit fraud

Honest, not open, not trustworthy
Keeping a type 3 secret mostly private, but sharing it discreetly in an inappropriate way.

Trustworthy, not open, not honest
Telling lies to hide an important secret

Open, not honest, not trustworthy
Sharing lots about yourself in an unfiltered stream-of-consciousness type of way, without checking your statements for accuracy or communicating clearly about the level of accuracy/carefulness of the statements

Open and Trustworthy, not Honest
This seems like the least common situation, but it is possible. One example would be sharing lots about yourself without carefully checking its accuracy, as a strategy to distract somebody from discovering a secret that you have been entrusted with.

Open and Honest, not Trustworthy
Blabbing a type 3 secret to everybody.

Trustworthy and Honest, not Open
General example: accurately sharing a type 3 secret to a particular person or in a particular fashion in which you are allowed to share it or asked to share it by the person whose secret it is.

Business example: a bank processing a credit card payment.

Open, Honest, and Trustworthy
Tactfully sharing your type 2 secrets with a friend who wants to hear them

Closing Thoughts

Since December 2021, I've been gradually shifting my personal values around the topics in this essay. I don't know how big the shift will be nor where all the chips will fall. It's part of my ongoing process of self-reflection. I think I will end up staying close to where I was before, still largely valuing honesty and openness and often trying to avoid generating and keeping secrets, but with more recognition of secret-keeping as an important skill/trait, one that is worth trying to get better at, and one that sometimes requires dishonesty and obfuscation (anti-openness). 

I will also be more aware that other people may understand the concepts of honesty, openness, and trustworthiness differently from me, and may choose a different balance in their lives with regard to these values. I will be less likely to harshly judge others' choices in that regard, as there are complex tradeoffs between objectives and virtues that are reasonably valued differently by different individuals.

  1. ^

    From a more sociological framework, I developed these values through my upbringing, they were instilled in me to an extent by my parents, especially my father, although I gave them my own interpretation. My father would often talk about how my grandfather (who I am named after) was a man of integrity, a man of his word who was trusted by his friends and business associates. 

  2. ^

    However, many secrets are mixtures of these types or do not fall into any of these categories. 

  3. ^

    The situation gets more complicated if two people in a friendship/relationship do not want equal levels of openness.

  4. ^

    This is oversimplified. I might also encourage you to share your type 1 secrets with me by exercising appropriate discretion in how I share my type 3 secrets with you and which ones I share, rather than hiding them completely. This might work if you do not want/expect others to keep all your type 1 secrets in an absolute sense. 

    Another exception might be if we are in some sense best friends or closest confidants in a committed way, so that there is an expected asymmetry, in line with social norms, in how I might share other people's secrets with you versus how I would share your secrets with others. 

    For example, there might be a social expectation that a married couple may ethically share most or all secrets between the two of them, including secrets that others have entrusted them with. For people who have not found a marriage partner or prefer not to pursue marriage or marriage-like relationships, there may not be an equivalent in this regard of a widely socially recognized closest confidant.

    Also, a patient/client/confessor may share secrets with a mental health professional/lawyer/ priest. Those are a very different context from the married-couple example as the secret-sharing relationship is not reciprocal.

  5. ^

    This is a bit of a stretch maybe, but I think it really is true in some cases. It's very context dependent. It is partly dependent on the cultural context of how or to what extent two of the other virtues (honesty and confidentiality) related to trustworthiness are viewed as aspects of the larger virtue of trustworthiness. 

    Of course, lying in a particular statement, almost by definition cannot signal honesty of one's communication in that particular statement, but it might signal that one is generally an honest person.

0 comments

Comments sorted by top scores.