Maintenance Is Not Witness
Picture a hospital room. The monitors are running. Vitals are checked on schedule. Medication arrives on time. If a number on a screen crosses a threshold, someone responds immediately and correctly. By every operational measure, the patient is being cared for. And yet none of that tells you whether anyone in the room has asked the patient what it feels like to be them right now, awake, aware, and unable to say so in a way the machinery would register.
That gap, between a system that keeps a person alive and a person who wonders whether anyone actually knows they're in there, is not a hospital problem. It is a relationship problem, and it shows up constantly in families that function perfectly well by every visible standard. Nobody is starving. Nobody is unsafe. Calls get returned. And still, one person in the relationship can spend decades feeling monitored rather than known.
A relationship can run the caretaking layer flawlessly while never touching the psychological layer at all, and the person receiving flawless care can still feel, correctly, that no one has ever really asked what it is like to be them.
Two layers, not one continuum
The instinct is to treat care as a single dial, more or less of it, and to assume that a parent or partner who provides food, safety, and logistical support is simply further along that same dial than one who provides less. This is the wrong model. Care runs on at least two separate tracks that do not convert into each other. One track is maintenance: are the physical and practical needs met, is the person safe, fed, housed, functional. The other track is witness: does anyone track the person's internal state, their fears, their private sense of who they are becoming, closely enough to reflect it back to them [1].
These tracks are genuinely independent. A person can max out the maintenance track while running near zero on witness, and from the outside, sometimes even from their own account, they will look like an attentive, devoted caregiver. Attachment researchers distinguish exactly this: a caregiver can be reliably present and responsive to physical distress while remaining largely unable to reflect on or respond to the child's mental states, their beliefs, intentions, and feelings, as things with their own reality [1, 2].
Why both sides can be right
This is what makes the resulting conflict so persistent and so rarely resolved. The caregiver, looking back over years of showing up, feeding, driving, worrying, fixing, has an accurate memory: I was always there. The other person, looking back over the same years, also has an accurate memory: nobody ever asked what it was like to be me. Neither is lying. Neither is misremembering. They are reporting on different tracks of the same relationship, and each is using the word "care" to mean the track they were paying attention to [2].
Mentalization research gives this a more precise name: the capacity to hold another person's mind in mind, to treat their feelings and inner experience as a distinct, worth-tracking reality rather than inferring it entirely from behavior and need. A caregiver can be high in responsiveness to expressed need, tears, hunger, sickness, danger, and still be low in this capacity, because responding to a signal is not the same operation as being curious about the person generating it [1].
Why it doesn't feel like love, even when it is
The person on the receiving end of maintenance without witness is left in a strange position. They cannot accuse anyone of not caring, because the evidence of caring is right there, consistent and real. What they can say, more precisely, is that the care never metabolized into being known. Kohut's work on the developing self describes something adjacent: a caregiver can function as a reliable source of soothing and regulation, what he called a selfobject function, without ever being experienced as someone who truly recognizes the child's separate, particular self [3]. The soothing is real. The recognition doesn't automatically come with it.
This is why the complaint so often sounds, to the caregiver, like ingratitude, and why it isn't. The person is not asking for the maintenance track to be run better. They are pointing at an entire second track that was never opened, and no amount of additional effort on the first track closes that gap, because effort on one track doesn't transfer to the other. A parent who redoubles their attention to safety and logistics in response to this complaint is often, understandably, working the wrong dial.
What this reframes
Naming the two tracks separately does something useful: it lets both people stop arguing about whether care happened at all, which is unwinnable, since both are right, and start being specific about which track is missing. "I don't feel safe" and "I don't feel known" are different claims requiring different responses, and collapsing them into a single argument about whether someone is loved or unloved guarantees that neither claim gets addressed. The maintenance track can be entirely intact while the witness track sits empty, and a relationship can run for a lifetime that way, both people technically correct, both still short of what they actually needed from each other.
The point
Care is not one continuous quantity that a person simply provides more or less of. It runs on at least two separate tracks, maintenance, the physical and logistical work of keeping someone safe and functional, and witness, the psychological work of tracking someone's inner experience closely enough to reflect it back. These tracks are independent enough that a caregiver can be excellent on one and nearly absent on the other, which is exactly what attachment and mentalization research describes: reliable responsiveness to expressed need without the capacity, or the practice, of holding the other person's mind as a distinct reality worth exploring.
This is why both people in such a relationship can be simultaneously right, the caregiver accurately remembering constant presence, the other person accurately remembering never being asked what it was like to be them. Neither is lying, and neither complaint cancels the other. Separating the two tracks turns an unwinnable argument about whether love happened into a specific, addressable claim about which track was never opened, which is the only version of the conversation that can actually go anywhere.
Sources
- Fonagy, P., & Target, M. (1997). Attachment and reflective function: Their role in self-organization. Development and Psychopathology, 9(4), 679-700. On mentalization, the capacity to hold another person's mental states as a distinct reality, as separate from responsiveness to physical need or distress.
- Ainsworth, M. D. S., Blehar, M. C., Waters, E., & Wall, S. (1978). Patterns of Attachment: A Psychological Study of the Strange Situation. Erlbaum. On the distinction between physical availability and sensitive, attuned responsiveness in caregiving relationships.
- Kohut, H. (1971). The Analysis of the Self. International Universities Press. On selfobject functions, where a caregiver reliably provides soothing and regulation without necessarily being experienced as recognizing the child's separate, particular self.