# Impama zothando
## A relationship snapshot of Mzilikazi and iKhohlwa lakhe
14 April–13 October 2026 · Compiled 14 October 2026

This portrait examines one WhatsApp export alongside published relationship research. It describes observable communication and several plausible interpretations. The user explains that much of the relationship is lived together and communicated through calls. That context is part of the analysis, not something to infer from texting gaps.

**Our own little vault:** the pictures, videos and voice notes omitted from the export are deliberately kept private between the two people, away from the machines. Their contents have not been analysed or reconstructed. Their omission is not interpreted as missing affection or insufficient sharing.

## 1. The snapshot

### A relationship lived in the everyday
The strongest repeated evidence is reciprocal affection and practical interdependence: meals, work check-ins, logistics, shared humour and plans for a home.

### The recurring need is clearer consultation
The most consequential visible tension concerns being informed, having time protected, and feeling that concerns lead to action. It appears across June, July and August.

### Repair exists; its durability is less observable
Both apologise or reassure. Some exchanges restore warmth. The written record is less complete about how new agreements are implemented in shared life.

### The quieter later chat is not a verdict
September and October contain fewer written messages per day after the documented move. Shared time, calls, changed routines and private media are plausible context. The record does not establish relationship decline.

The most defensible overall description is **reciprocal affection and practical interdependence, with repeated negotiation around consultation, attention and follow-through**. The record is rich in humour and ordinary care. It also contains direct complaints about communication and boundaries. Neither the affectionate moments nor the disagreements alone provide an impartial picture.

Visible written text is almost evenly divided. iKhohlwa starts more export-defined sessions across every tested time-gap threshold. Mzilikazi sends more omitted-media entries and somewhat more written word tokens. These differences describe the archive; they do not rank investment or devotion.

Warm exchanges later in the record are evidence of connection at those moments. Earlier requests about consultation and protected time remain meaningful because the written record does not demonstrate their consistent implementation. A conversation may have happened offline; absence from the export is not evidence that it did not.

## 2. What can be counted

| Measure | Mzilikazi | iKhohlwa lakhe | Total |
| --- | --- | --- | --- |
| All participant entries | 6001 | 5186 | 11187 |
| Inspectable text entries | 4146 | 4163 | 8309 |
| Private media entries | 1685 | 705 | 2390 |
| Blank entries | 163 | 276 | 439 |
| Deletion markers | 7 | 42 | 49 |
| Approximate word tokens | 24571 | 18965 | 43536 |
| URL occurrences | 66 | 332 | 398 |
| Messages with 😂 or 🤣 | 732 | 281 | 1013 |

The observation window has **183 calendar days**, with a participant entry on **182 days**. The only date with no exported entry is **11 October**. One encryption system notice is excluded. The source contains **five backward timestamp pairs**, corrected by stable chronological sorting before time-based calculations.

### Month by month

| Month | Mzilikazi | iKhohlwa | Total entries | Visible text | Private media | Observed days | Entries/day |
| --- | --- | --- | --- | --- | --- | --- | --- |
| 2026-04 | 475 | 398 | 873 | 639 | 205 | 17 | 51.35 |
| 2026-05 | 875 | 872 | 1747 | 1368 | 277 | 31 | 56.35 |
| 2026-06 | 1268 | 1003 | 2271 | 1487 | 682 | 30 | 75.7 |
| 2026-07 | 1255 | 1079 | 2334 | 1834 | 380 | 31 | 75.29 |
| 2026-08 | 1145 | 1148 | 2293 | 1844 | 373 | 31 | 73.97 |
| 2026-09 | 580 | 592 | 1172 | 995 | 138 | 30 | 39.07 |
| 2026-10 | 403 | 94 | 497 | 142 | 335 | 13 | 38.23 |

April covers 17 observed calendar days; October covers 13. Dividing by those observed days avoids treating the partial months as full months. September and October have fewer entries per day than June–August. The documented move occurs in September; more shared time, different routines, calls and private media are plausible context. A decline in relationship quality is not established.

The busiest exported day, **10 October**, has 249 entries: 246 from Mzilikazi and three from iKhohlwa. **236 are private media entries**. On **17 June**, 161 of 240 entries are private media. These examples show why total-message bars can be misleading if interpreted as equivalent conversational effort. The export does not identify what those privately kept moments contain.

### Who reopens the written channel?

| Gap threshold | Sessions started by Mzilikazi | Sessions started by iKhohlwa | Total |
| --- | --- | --- | --- |
| 2 hours | 204 | 335 | 539 |
| 4 hours | 98 | 171 | 269 |
| 6 hours | 68 | 111 | 179 |
| 8 hours | 46 | 77 | 123 |

A session starts at the first participant entry or after an interval **at least as long as** the chosen threshold without any participant entry. This is a descriptive convention, not a natural boundary between conversations. At six hours, starts are 68 versus 111. The direction survives the two-, four- and eight-hour definitions, so the finding is not an artefact of selecting six hours alone. Wake times, work schedules, calls and co-presence still prevent interpretation as total initiative.

The first qualifying explicit morning greeting occurs on only 32 dates: Mzilikazi 18, iKhohlwa 14. The rule requires a message of at most 120 characters containing “good morning” or “sawubona” between 04:00 and 11:59. This narrow subset misses many greeting conventions and should not overturn the broader session result. No greeting obligation is inferred.

### Writing bursts and adjacent reply gaps

| Measure | Mzilikazi | iKhohlwa |
| --- | --- | --- |
| Writing bursts | 3053 | 3256 |
| All adjacent sender changes | 2869 | 2869 |
| Sender changes within two hours | 2758 | 2668 |
| Median gap within that subset | 1.0 min | 2.0 min |
| Same-minute changes | 692 | 657 |

Writing bursts restart at a sender change or a gap of at least 30 minutes. The adjacent-gap calculation uses the last entry before a sender change, not the first message in a burst. A one- or two-minute median is **not read-to-reply time**. Timestamps have minute resolution; same-minute events are recorded as zero. The two-hour restriction removes overnight and other long gaps but does not establish availability or attention.

### Selected language cues

| Literal cue | Mzilikazi | iKhohlwa |
| --- | --- | --- |
| Laughing emoji | 732 | 281 |
| Heart symbol | 4 | 28 |
| English love-you phrase | 27 | 12 |
| English miss-you phrase | 21 | 4 |
| English thanks | 44 | 41 |
| Selected Nguni love forms | 1 | 1 |
| Selected Nguni thanks forms | 2 | 24 |
| Hug symbol | 49 | 10 |

Each count is the number of inspectable-text entries containing at least one specified match, not the number of symbols or words repeated inside that entry. English phrase matching permits “love you”, “miss you”, “thanks” or “thank you”; selected Nguni forms are deliberately incomplete. Elongated spellings and other affectionate forms may not match. The exact regular expressions are in the reproducibility script.

The two people use different written expressive cues. This supports describing different textures of expression, not stronger or weaker feelings. Literal English thanks is close in count, while the selected Nguni thanks forms appear more frequently in her text. Cultural and linguistic variation in verbal gratitude makes a universal gratitude score inappropriate [25]. The messages also include humour, requests and copied snippets that complicate a simple positive/negative sentiment classification.

Approximate word tokens use Unicode alphabetic runs, optionally joined by an apostrophe. URLs and edit markers are removed; media, blanks and deletion markers contribute zero. Some copied text remains in word and phrase totals. The result measures written lexical volume, not speech, disclosure depth or emotional competence.

## 3. Fourteen relationship themes

The categories below are qualitative lenses, developed from recurring content and contextual sequences. They are not exhaustive validated psychological constructs. Presence is easier to establish than frequency, intensity or a partner’s experience of it. The eight counted keyword groups used for the website’s constellation are a separate, explicitly mechanical analysis.

### 1. Chosen, and said out loud
**Classification:** Observed strength

Tenderness is repeated and reciprocal, in ordinary days as well as big declarations.

**Observed:** Both express love, longing and admiration. The 21 September exchange explicitly describes a positive impact on both lives.

**Alternative or limit:** A declaration shows what was said then; it does not measure satisfaction across all situations.

**A question to share:** Which ordinary gesture makes you feel most unmistakably chosen?

**Evidence:** 21 Sep — export entries 10405, 10409, 10419, 10421; 25 Jun — export entries 4548, 4549; 8 Aug — export entries 7807, 7809, 7826. Research lenses: [1], [2], [17].

### 2. Our own ridiculous little world
**Classification:** Observed strength

Jokes, teasing and playful exaggeration make the written exchange recognisable as theirs.

**Observed:** Bodily humour, food jokes, affectionate mock bargaining and shared reactions recur. Laughing emoji appear in 1,013 messages.

**Alternative or limit:** Emoji may soften teasing, mark embarrassment or accompany a forwarded clip; they are not proof that both people laughed.

**A question to share:** Which running joke still catches you off guard?

**Evidence:** 29 May — export entries 2410, 2412; 27 Aug — export entries 9319, 9321, 9322; 21 Sep — export entries 10406, 10410, 10411. Research lenses: [10], [12].

### 3. Come sit with me
**Classification:** Observed strength

Much of the chat helps two people who are already nearby find each other.

**Observed:** Invitations downstairs, meal breaks, bedtime, hugs and shower coordination show how texting serves shared space.

**Alternative or limit:** Same-house messages do not establish total hours together. Pictures, videos and voice notes are intentionally kept private; a quiet text day could reflect more shared time offline.

**A question to share:** What small pocket of uninterrupted time would you like more of?

**Evidence:** 15 Apr — export entries 48; 5 May — export entries 1174, 1176; 13 Jul — export entries 5798, 5802; 4 Sep — export entries 9689. Research lenses: [11], [18], [19].

### 4. Love, served warm
**Classification:** Observed strength

Food is a recurring language of practical care, pleasure and attention.

**Observed:** iKhohlwa prepares meals and checks preferences; Mzilikazi cooks, offers food and arranges alternatives when she is unwell. Both express appreciation.

**Alternative or limit:** Written visibility is uneven. These examples cannot tell us the true workload or whether each arrangement feels fair.

**A question to share:** When does making a meal feel loving, and when does it feel like another job?

**Evidence:** 8 May — export entries 1334, 1336; 10 Jun — export entries 3270, 3284; 15 Aug — export entries 8401, 8403; 1 Sep — export entries 9536, 9537. Research lenses: [5], [13], [14], [15].

### 5. The invisible little jobs
**Classification:** Needs negotiation

Care includes noticing, remembering and coordinating—not just doing the visible task.

**Observed:** Laundry, supplies, packing, utilities and cleaner instructions recur. She also follows up on unfinished tasks. He takes on laundry, cleaning and preparation in several examples.

**Alternative or limit:** Requests in text overrepresent coordination and miss completed work. No full domestic time-use record exists.

**A question to share:** Which recurring task could one person own from noticing it to finishing it?

**Evidence:** 12 May — export entries 1498; 5 Aug — export entries 7655, 7656, 7657; 12 Aug — export entries 8033, 8034, 8040; 20 Aug — export entries 8752, 8753. Research lenses: [13], [14], [15], [26].

### 6. Have you rested, mntuwam?
**Classification:** Observed strength

Checking on tiredness, comfort and work strain is a persistent form of attentiveness.

**Observed:** She repeatedly encourages rest during intense work periods. He offers comfort and practical help when she is unwell and checks how she is doing.

**Alternative or limit:** Health descriptions are self-reports. The export cannot establish medical outcomes or whether advice was useful.

**A question to share:** When you feel depleted, do you want comfort, a practical plan, company or quiet?

**Evidence:** 17 May — export entries 1623; 20 May — export entries 1879, 1880, 1883; 18 Jun — export entries 3991; 12 Sep — export entries 10124. Research lenses: [1], [5], [26].

### 7. I vouch for you
**Classification:** Observed strength

They show interest in each other’s work, ideas and achievements.

**Observed:** She encourages his projects and celebrates a new opportunity. He gives feedback on her logo and proposes support for her business.

**Alternative or limit:** Support can be received differently than intended. Encouragement and actual goal progress are separate observations.

**A question to share:** What kind of encouragement helps you keep going without adding pressure?

**Evidence:** 21 Apr — export entries 218, 220; 16 Jul — export entries 5947, 5957; 22 Jul — export entries 6290, 6293, 6296; 23 Jul — export entries 6428, 6436. Research lenses: [6], [7], [16].

### 8. Making the month work
**Classification:** Read with context

Money appears as shared problem-solving, stress and practical support.

**Observed:** Transport, utilities, groceries, debt and housing costs are discussed. Both contribute in visible instances, including her offering money or food help in tight periods.

**Alternative or limit:** This is not a financial ledger. Omitted transfers, non-cash work and outside obligations prevent a valid contribution total or dependence score.

**A question to share:** Can we discuss limits and help in a way that leaves both people room to say no?

**Evidence:** 23 Jun — export entries 4354; 25 Jun — export entries 4468, 4471; 21 Jul — export entries 6227; 1 Sep — export entries 9536, 9537. Research lenses: [5], [14], [16], [28].

### 9. A place for us
**Classification:** Observed strength

Shared plans become concrete through house hunting, moving and imagining a future.

**Observed:** They compare homes, discuss requirements and costs, organise viewings and later pack and move. Marriage and future family language recur.

**Alternative or limit:** Plans are evidence of intention; they do not guarantee shared expectations about roles, location or marriage.

**A question to share:** Which part of our imagined future is already agreed, and which part needs a fresh conversation?

**Evidence:** 24 Apr — export entries 504, 512, 513; 11 Jun — export entries 3347; 12 Aug — export entries 8055, 8056; 15 Sep — export entries 10241, 10280, 10302. Research lenses: [16], [18], [19].

### 10. A life bigger than two
**Classification:** Read with context

Family and parenting obligations shape time, availability and plans.

**Observed:** Childcare, school transport, relatives’ needs and family contact enter the shared logistics. He offers help; she sometimes declines it or explains competing demands.

**Alternative or limit:** The full family system is absent. Private third-party stories are not evidence of anyone’s character.

**A question to share:** How can support for family coexist with time and independence for us?

**Evidence:** 19 May — export entries 1828; 19 Jun — export entries 4041; 22 Jul — export entries 6356, 6357; 31 Jul — export entries 7159, 7164. Research lenses: [5], [16], [18].

### 11. Tell me before the plan changes
**Classification:** Needs negotiation

Consultation is a repeated relationship need, expressed by both people in different situations.

**Observed:** On 20 June she asks for earlier, direct communication about visits. He accepts responsibility. On 28 July he asks to hear changes before public announcements; she explains needing time to gather herself.

**Alternative or limit:** Neither perspective establishes the other’s intent. A request for information should still allow space, privacy and independent choices.

**A question to share:** What needs a heads-up, what needs a joint decision, and what is each person’s own choice?

**Evidence:** 20 Jun — export entries 4209, 4215, 4220, 4221; 28 Jul — export entries 6839, 6842, 6844, 6846, 6847. Research lenses: [3], [16], [22], [28].

### 12. Time that feels like ours
**Classification:** Needs negotiation

Affection sits alongside a clear wish for protected time and workable boundaries.

**Observed:** Call arrangements appear in May and June, then become explicit on 13 August. She describes an interruption and asks to discuss physical events in person. He proposes call windows and later acknowledges her point.

**Alternative or limit:** A proposed schedule and an apology are visible; consistent implementation of the schedule is not. Another partner’s comfort cannot establish hers.

**A question to share:** Which moments should be uninterrupted, and how do we agree exceptions?

**Evidence:** 29 May — export entries 2395, 2400; 15 Jun — export entries 3607; 13 Aug — export entries 8325, 8331, 8334, 8335, 8336; 15 Aug — export entries 8450, 8454, 8456. Research lenses: [3], [16], [23], [28].

### 13. Finding the way back
**Classification:** Read with context

They can apologise, offer reassurance and resume warmth; follow-through is less visible.

**Observed:** His 20 June apology identifies a specific omission. She apologises for being mean on 27 July. On 13 August he acknowledges her complaint and proposes talking in person.

**Alternative or limit:** Warmth afterward may coexist with unresolved practical issues. Calls or face-to-face repair may have happened outside the export.

**A question to share:** After an apology, what small action tells us the issue is being handled?

**Evidence:** 20 Jun — export entries 4220, 4221, 4222, 4225; 27 Jul — export entries 6669, 6670; 13 Aug — export entries 8336; 13 Sep — export entries 10156, 10157. Research lenses: [3], [4], [27].

### 14. Attraction, with a voice
**Classification:** Read with context

Desire and physical closeness are openly expressed, with preferences and limits also spoken.

**Observed:** The record includes invitations, longing, affectionate touch and statements about preferred or unwanted timing. The August call dispute also concerns feeling interrupted during closeness.

**Alternative or limit:** Flirtation cannot establish overall sexual satisfaction or continuing consent. A September nightmare is a reported dream, not an observed act.

**A question to share:** How do we check that both of us still want the same thing, in the moment?

**Evidence:** 18 May — export entries 1688; 22 Jun — export entries 4348; 13 Aug — export entries 8334; 12 Sep — export entries 10127. Research lenses: [1], [16], [24].

## 4. Six sequences in context

### 17–27 MAY — Care when work fills the room
**State visible in the record:** Visible support

**Concern:** Work pressure and exhaustion repeatedly enter the conversation.

**Response:** iKhohlwa encourages rest, food and hydration; Mzilikazi makes room for a call even while busy.

**Afterward:** He later describes feeling well rested. A causal link to any single message cannot be established.

**Reading:** Being interested in the person behind the work is a recurring strength.

Evidence entries: 1623, 1816, 1879, 1883, 2380. Research lenses: [1], [5], [6].

### 20 JUNE — An apology that names the omission
**State visible in the record:** Acknowledged; follow-through unseen

**Concern:** A visit is mentioned after it has already happened. iKhohlwa asks why she was not consulted earlier.

**Response:** Mzilikazi explicitly accepts that he failed to communicate and says she has not made communication difficult.

**Afterward:** The next day opens with a warm greeting. The topic later recurs around calls and protected time.

**Reading:** Accountability is visible. A lasting new routine is still an open question.

Evidence entries: 4206, 4209, 4215, 4220, 4221, 4227, 4228. Research lenses: [3], [4], [27].

### 21–30 JULY — Two versions of what needs saying
**State visible in the record:** Negotiation remains visible

**Concern:** Work changes, housing costs, visit timing and personal plans create a complicated week.

**Response:** She asks for more discussion and explains needing time to gather herself. He asks to hear important changes before public updates.

**Afterward:** Plans to meet and continued house hunting coexist with tension. The transcript does not document a full shared rule for consultation.

**Reading:** The shared question is how to inform each other without erasing either person’s independence.

Evidence entries: 6227, 6784, 6816, 6842, 6844, 6847, 7114. Research lenses: [3], [16], [22].

### 13 AUGUST — A call window is only part of the answer
**State visible in the record:** Promised in-person conversation

**Concern:** Mzilikazi proposes predictable call windows while another partner is visiting.

**Response:** iKhohlwa describes an interruption during closeness and says the issue includes how events are repaired, not just when calls happen.

**Afterward:** He says she is right, apologises and proposes speaking in person the following day. The export does not show that conversation.

**Reading:** A timetable may help predictability. It also needs an agreement about attention, privacy and in-person repair.

Evidence entries: 8325, 8331, 8333, 8334, 8335, 8336. Research lenses: [3], [16], [24], [28].

### 13 SEPTEMBER — A dream, and a reassurance
**State visible in the record:** Reassurance documented

**Concern:** iKhohlwa reports a distressing nightmare involving changed agreements and an ignored request.

**Response:** She asks if it will happen. Mzilikazi reassures her about the stated relationship limit and respecting a request to take a call outside.

**Afterward:** Ordinary affectionate and practical conversation continues. The nightmare does not verify that its events happened.

**Reading:** The reassurance is observable; an attachment diagnosis or allegation based on dream content would be unsupported.

Evidence entries: 10137, 10148, 10156, 10157. Research lenses: [1], [16], [22].

### 21 SEPTEMBER — Just love me like this
**State visible in the record:** Mutual appreciation

**Concern:** Mzilikazi asks how to express how much she has enriched his life.

**Response:** iKhohlwa answers that he makes it easy to love him and asks him to keep loving her as he already does.

**Afterward:** Both describe a positive impact on their lives, with humour woven through the tenderness.

**Reading:** This is a strong example of appreciation flowing in both directions. It coexists with earlier needs for clearer agreements.

Evidence entries: 10405, 10409, 10419, 10421. Research lenses: [8], [9], [15].

## 5. Two observable ways of showing up

### Mzilikazi
The record shows expressive affection, practical support, willingness to share detailed plans, interest in what she thinks and capacity for specific accountability. On 20 June he identifies his failure to mention a visit, recognises that delaying information made things harder and explicitly says she has not made communication difficult. In August he proposes a practical call arrangement, then acknowledges that her concern is broader than the timetable.

The recurring pressure point is communicating after a situation has already developed. Structured explanations can clarify an issue but can also answer logistics before recognising what the other person experienced. This is an interactional observation rather than a fixed temperament or diagnosis. His greater media volume also means that text alone may disproportionately display the explanatory side of his communication.

### iKhohlwa lakhe
The record shows attentive practical care, encouragement, humour, organisation and a direct willingness to ask for common ground. Her June and August messages clearly name what is not working rather than requiring the reader to infer it. She expresses preferences, proposes alternatives and sometimes declines assistance. She supports projects and describes his positive impact on her life.

The recurring pressure point is making a pause or changing plan understandable to him while preserving her own space. On 28 July she explains that she wanted time to gather herself before giving him a fuller account. A wish for space and a wish for a timely heads-up can both be understandable. No global conclusion about evasiveness, jealousy or instability is supported by that sequence.

### The relationship between the styles
The strongest common ground is not identical expression. It is a demonstrated ability to contribute, express love, ask for the other’s attention and negotiate parts of shared life. The useful question is whether both people can turn those negotiations into arrangements that feel clear, respectful and workable. Neither another partner’s preferences nor a researcher’s sample can substitute for their own agreement.

## 6. The constellation: a map of word co-occurrence

| Keyword group | Mzilikazi matches | iKhohlwa matches | Total |
| --- | --- | --- | --- |
| Love words | 613 | 383 | 996 |
| Food words | 109 | 110 | 219 |
| Work words | 109 | 39 | 148 |
| Home words | 105 | 140 | 245 |
| Money words | 73 | 54 | 127 |
| Family words | 29 | 35 | 64 |
| Health words | 88 | 103 | 191 |
| Planning words | 89 | 79 | 168 |

A node count is the number of inspectable messages matching that group’s explicit regular expression. A message can belong to several groups. A connection counts messages matching both groups; it is **not** a correlation, a causal arrow, a direction of support or a compatibility score. The graph emphasises the 12 largest connections initially; selecting a node highlights all its connections.

The rules include common English words and a limited selection of Nguni forms. “Water” can mean hydration or household supply; “baby” can be an endearment or a child; “home” can refer to either household. Those ambiguities are why the nodes are called **word groups**, rather than claims about eight quantified emotional states. Known copied passages and automated payment/utility notices are excluded from this graph by an explicit list and prefix rules. This is not a complete automatic forwarding detector.

### Love words
`\b(?:love|lovey|lovery|baby|sthandwa|mamakhe|mamazi|lovey\s+dove)\b|ngikuthand|ndiyakuthand|ngiyakuthand`

### Food words
`\b(?:food|meal|eat|eating|breakfast|lunch|supper|dinner|cook|cooking|tea|coffee|bread|chicken|rice|braai|salad)\b|upheka|kupheka|ndipheke`

### Work words
`\b(?:work|working|meeting|project|business|job|gig|logo|website|salary)\b|umsebenzi|emsebenz|msebenzi`

### Home words
`\b(?:house|home|bed|kitchen|clean|cleaning|laundry|washing|clothes|windows|blanket|blankets|geyser|shower)\b|endlini|ekhaya|impahla|izitya`

### Money words
`\b(?:money|rand|rent|pay|paid|payment|bank|deposit|loan|bills|budget|electricity)\b|imali|umbane|\br\s?\d+`

### Family words
`\b(?:child|kid|kids|toddler|school|mom|mother|dad|father|sister|brother|parents|family)\b|umntana|umntwana|eskolweni|isikolo|umntanam`

### Health words
`\b(?:sick|pain|sleep|slept|rest|tired|hydrate|water|clinic|medication|medicine|meds|cramps|headache|backache|flu)\b|ndidiniwe|ndigodola|ndilala|ulala|ulale`

### Planning words
`\b(?:plan|plans|tomorrow|weekend|trip|flight|travel|moving|viewing|book|booking|married|marry|wedding)\b|ngomso|kusasa|ksasa`

### All counted connections

| Connection | Messages matching both |
| --- | --- |
| Home words × Love words | 31 |
| Food words × Love words | 26 |
| Health words × Love words | 23 |
| Love words × Money words | 18 |
| Love words × Work words | 17 |
| Health words × Planning words | 15 |
| Love words × Planning words | 15 |
| Health words × Home words | 15 |
| Food words × Home words | 13 |
| Food words × Health words | 12 |
| Health words × Work words | 11 |
| Home words × Planning words | 11 |
| Planning words × Work words | 10 |
| Health words × Money words | 8 |
| Food words × Work words | 8 |
| Home words × Money words | 8 |
| Family words × Planning words | 7 |
| Home words × Work words | 7 |
| Food words × Money words | 6 |
| Money words × Planning words | 5 |
| Food words × Planning words | 5 |
| Money words × Work words | 4 |
| Family words × Money words | 3 |
| Family words × Work words | 3 |
| Family words × Home words | 2 |
| Family words × Love words | 2 |
| Family words × Health words | 2 |

## 7. Questions this snapshot makes worthwhile

1. What made you feel cared for this week that I might not have noticed?

2. Would you like comfort, company, practical help or space right now?

3. Which plan needs a heads-up before it changes?

4. What should stay uninterrupted when we are together?

5. Which ordinary task could I take from start to finish?

6. What would make an apology feel complete to you?

7. What is one little adventure we would both enjoy?

8. What do you want me to keep doing exactly as I already do?

These are conversation prompts developed from the observations, not a therapeutic intervention that has been tested on this couple. The practical emphasis is on checking what support is wanted, defining which plans require consultation, protecting time and making follow-through visible.

A potentially useful distinction is **a heads-up versus a joint decision versus a personal choice**. Another is **an apology versus a revised agreement versus evidence that it is working**. Neither distinction awards authority to one person; both invite an explicit shared understanding.

## 8. Method, context and limits

**Inputs and units.** One attached WhatsApp ZIP, containing one exported text file, was parsed into participant entries. Multiline content is preserved. Entry identifiers in this report are one-based positions in the original participant-entry order, before timestamp sorting. They allow the owner to inspect context without republishing the transcript. The date range is inclusive. Local South African time is assumed because the export contains no timezone offset.

**Reading workflow.** The whole export underwent a computational scan for structure and lexical cues. Longer conversational messages were reviewed chronologically, with surrounding short turns and targeted searches used for close reading of the selected sequences. This does not mean every entry received independent manual semantic coding. Omitted media were intentionally left private. Third-party stories were not treated as verified allegations, and quoted/copied snippets were not automatically attributed to the sender’s personal experience. In particular, the 13 September account is explicitly introduced as a nightmare.

**Qualitative selection.** Themes were developed from recurrence and substantive relevance. Six case studies were selected to include support, appreciation, consultation, boundaries, repair and reassurance. Selection can overrepresent issues that are long enough to discuss in text. The cases are illustrative rather than a census of every disagreement or affectionate exchange. There is no independent second coder or inter-rater reliability estimate. Labels describe observable patterns, not validated psychological categories.

**Missing observation versus chosen privacy.** Calls and face-to-face life are outside the export. Private media are deliberately excluded, not a failure to share. Blank entries and deletion markers cannot be reconstructed. The visible corpus is therefore selective in several different ways. Neither message totals nor silence can stand in for a complete relationship.

**Dependent data.** Thousands of messages from one couple are not thousands of independent couples. No significance test is used to infer population-level relationship quality. No sentiment model, clinical attachment classification, emotional-intelligence score, health score, abuse diagnosis, satisfaction questionnaire or breakup prediction is produced.

**Research approach.** This is a focused interpretive literature review, not a systematic review. Searches covered responsiveness, gratitude, positive-event sharing, laughter, conflict recovery, household cognition and fairness, autonomy, texting, language, cultural generalisability and multi-partner maintenance. All 28 references were consulted through bibliographic records/abstracts; full text was additionally consulted for entries 12, 14 and 17. Publisher or PubMed access limits affected some other page opens. Access level is recorded for every entry. Meta-analyses, diary studies, experiments and reviews provide different kinds of evidence. They are not interchangeable.

**Transporting evidence.** Most cited samples do not closely match this couple’s cultural, linguistic or family setting. Conventional dyadic research and consensual non-monogamy evidence provide qualified questions; they do not determine what customary arrangements should be. Source 21 addresses generalisation; source 23 addresses monogamy assumptions. Language-style matching research [20] motivates measurement caution, but no language-style matching score is calculated. Source 22 distinguishes knowing someone’s inner state from the appearance of understanding in a transcript.

**Discretion.** The published site and report contain aggregates, paraphrased sequences and six short affectionate excerpts. They exclude the raw export, contact information, exact addresses and locations, financial redemption codes, emergency phrases, private health specifics and third-party allegations. The little vault is a description of your privacy boundary, not an online repository of private media.

**Checking the result.** The numerical results can be reproduced by passing the original export to `scripts/analyse.py`. The build script produces the report, data bundle and standalone preview. Validation reconciles participant, monthly, daily and category totals and checks source entry identifiers. The website presents qualitative findings separately from literal keyword counts.

## 9. Design lineage

The earlier **Good at Emotion** package is explicitly an iBibi lam relationship site. Its HTML, JavaScript, stylesheet, report structure and supplied 390-pixel mobile preview were inspected. No live deployed URL was established from the available references, and no separate second iBibi website was located. The review therefore uses the actual supplied source and design preview, not an invented live deployment.

That site uses a pastel research-dashboard language: a large introductory hero, coloured metric cards, charts, 12 thematic cards, case studies, two portraits, a research section and an expanded report. Impama zothando uses a different visual structure: an ivory keepsake, rose ink, sparse gold lines, a wax-sealed letter, arch-shaped petals, a dark rose chapter for nuanced sequences, a private-vault dedication and thumb-friendly chapter navigation.

The narrative leads with lived meaning, then offers measurable rhythm, themes, contextual sequences, two perspectives and research. Individual observations, alternative interpretations and evidence links are reachable from each thematic petal. Charts use touch selection; case studies use native horizontal swipe scrolling with buttons and keyboard alternatives. Decorative tilt and vibration are optional enhancements, never prerequisites for reading.

## 10. Research reading list

### [1] Laurenceau, Barrett & Pietromonaco (1998)
**Intimacy as an interpersonal process**

Study type: Interaction diary studies. Access consulted: Abstract/index. Checked 14 October 2026.

Lens: Disclosure and feeling understood help explain intimacy.

Application limit: Messages do not measure either partner’s felt responsiveness.

[Open the primary research record](https://pubmed.ncbi.nlm.nih.gov/9599440/)

### [2] Laurenceau et al. (2005)
**The interpersonal process model of intimacy in marriage**

Study type: Couple diary study. Access consulted: Abstract/index. Checked 14 October 2026.

Lens: Intimacy can be studied as a repeated interpersonal process.

Application limit: Married samples and daily self-reports differ from this export.

[Open the primary research record](https://pubmed.ncbi.nlm.nih.gov/15982109/)

### [3] Gordon & Chen (2016)
**Do you get where I’m coming from? Perceived understanding buffers against the negative impact of conflict**

Study type: Seven studies, including experiments. Access consulted: Abstract/index. Checked 14 October 2026.

Lens: Feeling understood during conflict can matter beyond reaching a solution.

Application limit: This record contains no validated understanding ratings.

[Open the primary research record](https://pubmed.ncbi.nlm.nih.gov/26523997/)

### [4] Salvatore et al. (2011)
**Recovering from conflict in romantic relationships**

Study type: Observed conflict and follow-up. Access consulted: Abstract/index. Checked 14 October 2026.

Lens: Recovery after an argument is a distinct process worth examining.

Application limit: A return to affectionate texting does not establish resolution.

[Open the primary research record](https://pubmed.ncbi.nlm.nih.gov/21245491/)

### [5] Falconier, Jackson, Hilpert & Bodenmann (2015)
**Dyadic coping and relationship satisfaction: A meta-analysis**

Study type: Meta-analysis. Access consulted: Abstract/index. Checked 14 October 2026.

Lens: How partners cope with stress together is associated with satisfaction.

Application limit: Practical requests alone do not show whether help felt supportive.

[Open the primary research record](https://pubmed.ncbi.nlm.nih.gov/26295276/)

### [6] Gable et al. (2004)
**What do you do when things go right?**

Study type: Four studies. Access consulted: Abstract/index. Checked 14 October 2026.

Lens: Sharing good news and an engaged response can support connection.

Application limit: The tone of calls and in-person celebrations is unavailable.

[Open the primary research record](https://pubmed.ncbi.nlm.nih.gov/15301629/)

### [7] Gable, Gonzaga & Strachman (2006)
**Will you be there for me when things go right?**

Study type: Positive-event response studies. Access consulted: Abstract/index. Checked 14 October 2026.

Lens: Responses to a partner’s good news deserve their own attention.

Application limit: Encouraging messages are examples, not a validated response-style score.

[Open the primary research record](https://pubmed.ncbi.nlm.nih.gov/17059309/)

### [8] Algoe et al. (2013)
**The social functions of the emotion of gratitude via expression**

Study type: Laboratory couple study. Access consulted: Abstract/index. Checked 14 October 2026.

Lens: Expressed gratitude can contribute to relationship-building.

Application limit: Counting thanks does not measure felt appreciation.

[Open the primary research record](https://pubmed.ncbi.nlm.nih.gov/23731434/)

### [9] Algoe et al. (2022)
**Implementation intentions to express gratitude increase daily time co-present with an intimate partner**

Study type: Five-week randomized field experiment. Access consulted: Abstract/index. Checked 14 October 2026.

Lens: A deliberate gratitude practice increased shared time in this study.

Application limit: That intervention was not administered here; its result is not a forecast.

[Open the primary research record](https://pubmed.ncbi.nlm.nih.gov/35810173/)

### [10] Kurtz & Algoe (2015)
**Putting laughter in context: Shared laughter as behavioral indicator of relationship well-being**

Study type: Observed laughter in 71 couples. Access consulted: Abstract/index. Checked 14 October 2026.

Lens: Shared laughter was associated with closeness and support.

Application limit: Laughing emoji are not measured simultaneous audible laughter.

[Open the primary research record](https://pubmed.ncbi.nlm.nih.gov/26957946/)

### [11] Holtzman et al. (2021)
**Long-distance texting: Text messaging is linked with higher relationship satisfaction in long-distance relationships**

Study type: Survey comparing relationship distance. Access consulted: Abstract/index. Checked 14 October 2026.

Lens: Communication channel and distance change what texting can mean.

Application limit: This is not a long-distance-only relationship; observational associations are not causal.

[Open the primary research record](https://pmc.ncbi.nlm.nih.gov/articles/PMC8669216/)

### [12] Eun Huh (2025)
**The impact of emojis on perceived responsiveness and relationship satisfaction in text messaging**

Study type: Randomized message scenarios. Access consulted: Full text. Checked 14 October 2026.

Lens: Emoji can change perceptions of responsiveness.

Application limit: 260 participants imagined messages from friends, not their actual romantic partners.

[Open the primary research record](https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0326189)

### [13] Allison Daminger (2019)
**The Cognitive Dimension of Household Labor**

Study type: Qualitative household research. Access consulted: Abstract/index. Checked 14 October 2026.

Lens: Anticipating, planning and monitoring are part of domestic work.

Application limit: A text export cannot measure the complete division of labour.

[Open the primary research record](https://journals.sagepub.com/doi/10.1177/0003122419859007)

### [14] Gillespie, Peterson & Lever (2019)
**Gendered perceptions of fairness in housework and shared expenses**

Study type: Cross-sectional survey. Access consulted: Full text. Checked 14 October 2026.

Lens: Perceived fairness in chores and expenses relates to relationship outcomes.

Application limit: Self-selected survey responses do not establish causation or fairness in this couple.

[Open the primary research record](https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0214204)

### [15] Gordon et al. (2022)
**Feeling Appreciated Buffers Against the Negative Effects of Unequal Division of Household Labor**

Study type: Three samples during COVID-19. Access consulted: Abstract/index. Checked 14 October 2026.

Lens: Feeling appreciated moderated links between unequal work and satisfaction.

Application limit: Appreciation does not make unwanted or unmanageable responsibilities acceptable.

[Open the primary research record](https://pubmed.ncbi.nlm.nih.gov/35819837/)

### [16] Patrick, Knee, Canevello & Lonsbary (2007)
**The role of need fulfillment in relationship functioning and well-being**

Study type: Three studies; self-determination framework. Access consulted: Abstract/index. Checked 14 October 2026.

Lens: Autonomy, competence and relatedness provide lenses for care and choice.

Application limit: No need-fulfilment questionnaire was used here.

[Open the primary research record](https://pubmed.ncbi.nlm.nih.gov/17352602/)

### [17] Impett, Park & Muise (2024)
**Popular Psychology Through a Scientific Lens: Evaluating Love Languages**

Study type: Research review. Access consulted: Full text. Checked 14 October 2026.

Lens: Evidence does not justify assigning one fixed primary love language.

Application limit: Use multiple forms of care as descriptions rather than personality categories.

[Open the primary research record](https://journals.sagepub.com/doi/full/10.1177/09637214231217663)

### [18] Ogolsky & Bowers (2013)
**A meta-analytic review of relationship maintenance and its correlates**

Study type: Meta-analysis. Access consulted: Abstract/index. Checked 14 October 2026.

Lens: Openness, assurances, positivity, shared tasks and networks are maintenance behaviours.

Application limit: Correlations do not score this relationship or predict its duration.

[Open the primary research record](https://doi.org/10.1177/0265407512463338)

### [19] Aron et al. (2000)
**Couples’ shared participation in novel and arousing activities and experienced relationship quality**

Study type: Survey and experimental studies. Access consulted: Abstract/index. Checked 14 October 2026.

Lens: Shared novel activities offer a lens for outings and future adventures.

Application limit: Mentioning an outing does not show that it occurred or improved satisfaction.

[Open the primary research record](https://pubmed.ncbi.nlm.nih.gov/10707334/)

### [20] Ireland et al. (2011)
**Language style matching predicts relationship initiation and stability**

Study type: Natural-language observational studies. Access consulted: Abstract/index. Checked 14 October 2026.

Lens: Linguistic coordination can be studied, but requires defined measures.

Application limit: Code-switching and shared nicknames are not a validated language-style-matching score.

[Open the primary research record](https://pubmed.ncbi.nlm.nih.gov/21149854/)

### [21] Henrich, Heine & Norenzayan (2010)
**The weirdest people in the world?**

Study type: Cross-cultural methodological review. Access consulted: Abstract/index. Checked 14 October 2026.

Lens: Research samples can be poor guides to universal human behaviour.

Application limit: Western couple studies cannot silently become Nguni cultural norms.

[Open the primary research record](https://pubmed.ncbi.nlm.nih.gov/20550733/)

### [22] Sened et al. (2017)
**Empathic accuracy and relationship satisfaction: A meta-analytic review**

Study type: Meta-analysis. Access consulted: Abstract/index. Checked 14 October 2026.

Lens: Accurately knowing another’s inner state is distinct from trying to understand.

Application limit: We have no independent report of either person’s thoughts or feelings.

[Open the primary research record](https://pubmed.ncbi.nlm.nih.gov/28394141/)

### [23] Conley et al. (2017)
**Investigation of Consensually Nonmonogamous Relationships**

Study type: Research review. Access consulted: Abstract/index. Checked 14 October 2026.

Lens: Monogamy should not be treated as the automatic benchmark of functioning.

Application limit: CNM evidence is not interchangeable with South African customary polygyny.

[Open the primary research record](https://pubmed.ncbi.nlm.nih.gov/28346120/)

### [24] Allen B. Mallory (2022)
**Dimensions of couples’ sexual communication, relationship satisfaction, and sexual satisfaction**

Study type: Meta-analysis of 93 studies. Access consulted: Abstract/index. Checked 14 October 2026.

Lens: Quality of sexual communication had stronger associations than frequency.

Application limit: Explicit desire in messages does not establish satisfaction, safety or ongoing consent.

[Open the primary research record](https://pubmed.ncbi.nlm.nih.gov/34968095/)

### [25] Floyd et al. (2018)
**Universals and cultural diversity in the expression of gratitude**

Study type: Cross-cultural natural interaction study. Access consulted: Abstract/index. Checked 14 October 2026.

Lens: Verbal thanks varies across cultural and linguistic settings.

Application limit: English phrase totals are incomplete measures of gratitude.

[Open the primary research record](https://pubmed.ncbi.nlm.nih.gov/29892463/)

### [26] Bolger, Zuckerman & Kessler (2000)
**Invisible support and adjustment to stress**

Study type: Diary research around a major stressor. Access consulted: Abstract/index. Checked 14 October 2026.

Lens: Some useful support is not consciously registered as a support transaction.

Application limit: Unseen help cannot be counted or attributed from this record.

[Open the primary research record](https://pubmed.ncbi.nlm.nih.gov/11138764/)

### [27] Haydon et al. (2017)
**The Dyadic Construction of Romantic Conflict Recovery Sabotage**

Study type: Longitudinal observational study of 100 couples. Access consulted: Abstract/index. Checked 14 October 2026.

Lens: What happens after a disagreement is part of the interaction, too.

Application limit: This coding scheme was not applied; neither partner is labelled a saboteur.

[Open the primary research record](https://pubmed.ncbi.nlm.nih.gov/29398761/)

### [28] Multiple-author study (2026)
**How Do People Maintain Consensual Non-Monogamy? An International Development and Validation of the Multiple Relationships Maintenance Scale**

Study type: Scale development and validation. Access consulted: Abstract/index. Checked 14 October 2026.

Lens: Disclosure and thoughtful distribution of resources are relevant multi-partner questions.

Application limit: No scale was administered; CNM samples cannot settle customary arrangements.

[Open the primary research record](https://pubmed.ncbi.nlm.nih.gov/41840193/)