Land, Water and Living Stewardship

Executive summary

Spiralweb works with places that differ in climate, history, tenure, water, knowledge, livelihood and institutional setting. The common task is therefore not to distribute one regenerative model. It is to support a disciplined way of finding out what a place can carry, what its stewards want to protect or change, and what the consequences of a next step actually are.

Land and water are treated here as one living field. Water availability cannot be read apart from soil cover, vegetation, geology, extraction, pollution, labour, rights and public infrastructure. A greener field is not necessarily a more water-secure field. A technically efficient intervention can still increase total demand, shift harm elsewhere or become impossible to maintain. Conversely, a modest act of protection, repair or assisted natural regeneration may be more consequential than a visible planting programme.

The proposed Spiralweb contribution is a place-held learning architecture. It connects local observation, scientific and technical knowledge, institutional memory, resource flows and public learning without giving any one source automatic sovereignty. The field can correct the record. The steward can refuse. Unknown remains a valid state. Money does not buy tempo, authority or a favourable narrative.

The paper advances six working propositions.

• No single practice family is regenerative by definition. Agroforestry, food forests, syntropic systems, water harvesting, irrigation, grazing management, wells and restoration can each help, fail or displace harm depending on place and design.

• The strongest transferable unit is not a technique but an inquiry grammar: listen, bound the question, build a proportional baseline, compare options, make the smallest responsible movement, observe consequences, return and correct.

• Ecological improvement cannot compensate for exhausted stewards or illegitimate governance. Spiralweb therefore keeps Land and Life, Steward Viability, and Governance and Commons as separate readings.

• Scientific evidence should inform local judgment without pretending that results from one climate, land use or social setting transfer intact to another. Local knowledge is indispensable and also fallible. Both require provenance, dialogue and correction.

• A PG Ledger becomes real only when dated observations, decisions, resource flows and Returns alter later action. Until that cadence exists, Spiralweb has proposals, baselines and early formats rather than a validated operating system.

• The present field relationships are at different stages. Doukkala and Firmaty are living relationships with preparatory field inquiries. Kitgum has an early locally authored readiness record and one Dashboard reading, not a regular Ledger cadence. Xochimilco is a mature social-ecological field in which Spiralweb may support a bounded relationship, not introduce regeneration to an empty place.

One grammar, several names

Purpose and scope

This paper provides a common professional frame for Spiralweb’s work where land, water, food, habitat and human livelihood meet. It is meant to sit between the wider Spiralweb corpus and field-specific agreements. It clarifies what may be shared across fields, what must remain local, what evidence can presently support, and what must still be investigated.

It draws on the current Spiralweb websites, the PG Ledger and five operative formats, Healthy Roots, Knowing from the Ground, The Living Field, Living Landscapes, The Spiral of Hypotheses, Flow Desk material, the Conscious Co-Created Food Systems protocol, and current working records for Doukkala, Firmaty, Kitgum and Xochimilco. Older Morocco and food-forest documents are used as design history and research leads, not as evidence of current agreements.

The inquiry grammar was first tested in a Danish context in At holde jorden levende (September 2026). That paper is not an application of this one. This paper is an attempt to make the Danish inquiry grammar legible across different climates without making it universal.

What this paper does not do

• It does not select species, earthworks, irrigation, wells, grazing systems or restoration targets for any field.

• It does not state that a grant or field funding has been received.

• It does not convert a relationship, proposal, public horizon or indicative budget into an active project.

• It does not make Xochimilco, permaculture, food forests or syntropic agriculture a universal template.

• It does not treat a green status as certification, a red status as punishment or an unknown status as hidden failure.

• It does not replace legal, hydrogeological, agronomic, ecological, public health or financial due diligence.

Core definitions

Land and water as one field

Water is not an input that can be understood through supply alone. It falls, flows, infiltrates, evaporates, is stored, pumped, polluted, allocated and embodied in crops and livelihoods. Soil structure and organic matter affect infiltration and plant-available water. Vegetation can protect soil and support infiltration, yet trees also transpire water and may reduce catchment yield in some settings. Irrigation efficiency can reduce applied water per unit while total withdrawal or consumptive use still rises through expansion, changed crops or altered return flows. A deeper well can protect a livelihood in the short term and still contribute to aquifer decline. Each statement can be true at the same time.

The practical unit is therefore a water and land relation, not a technology. The first question is rarely “Which technology should we install?” It is “What water relation are we trying to protect or change, for whom, over what season and boundary, and what evidence would show improvement or harm?”

A living field can change how water moves

The ecological wager is not that planting automatically creates water. It is that persistent cover, roots, litter, soil organisms and added biomass may change how rainfall and irrigation are partitioned: less rapid runoff and bare-soil evaporation; more infiltration, moderated surface temperature, soil storage and productive transpiration. Roots and organic inputs may improve aggregation and pore structure, while shade, wind reduction and surface roughness can slow drying. These functions can make a field more able to receive and hold a pulse of water. They do not remove the need to account for where that water came from.

More soil organic matter is therefore relevant, but it is not a water-volume promise. The additional plant-available water associated with a deeper or more organic topsoil depends on texture, bulk density, rooting depth and the field’s water-retention curve; published synthesis finds the average direct effect of organic matter on available water capacity to be modest and variable (Minasny and McBratney 2018). A claim such as “ten centimetres more topsoil stores a given number of litres” must be calculated from measurements at the site, not from a universal conversion.

Trees add another scale. Canopy, litter and roots may increase infiltration, shade and microclimatic buffering, and transpired water participates in atmospheric moisture recycling. Trees also use water. Research in seasonally dry tropical systems found the highest groundwater recharge at intermediate rather than maximal tree cover (Ilstedt et al. 2016). A global synthesis of catchment studies found that afforesting grasslands or shrublands usually reduced annual runoff, with proportionally greater losses at drier sites (Farley et al. 2005). Neither study predicts the Moroccan fields. Together they require a diagnosis of the starting condition and the water pathway: degraded bare or crusted ground may lose rainfall as runoff, while adding dense woody cover to an already vegetated water-limited system may raise evapotranspiration and reduce downstream yield. The question is which living structure, at what density and over what boundary, improves the field without exporting an unseen water cost.

This gives the land inquiry a measurable direction. A proposed regenerative design should state which pathway it expects to change—cover, infiltration, erosion, evaporation, soil moisture, biomass cycling, crop survival, irrigation demand or groundwater dependence—and over what time. Reference areas and repeated observations can then show whether the field is becoming more water-capable, merely greener because more water is supplied, or less viable despite the intervention.

Two scales of correction

A field can correct what it can observe at field scale: cover, bare ground, irrigation applied, reservoir level, crop survival, labour, cost, soil moisture and visible erosion. These observations can improve the next local movement. They cannot by themselves determine long-term aquifer storage, recharge, salinity movement, downstream flow or cumulative basin use. Those questions need hydrogeological records, catchment monitoring, public data or specialist work at the scale and duration the decision requires.

External evidence does not become sovereign over lived knowledge, and local experience does not substitute for processes that cannot be sensed from one plot. The Ledger must hold both with their provenance and scale. Where the wider effect matters but remains unresolved, the affected-field standing stays Unknown, the claim is narrowed, or Spiralweb pauses its own support. Unknown need not be permanent, but it cannot be converted to Green because the local intervention appears successful.

The catchment is an outer standing

A steward-centred field boundary is necessary but insufficient for water. Aquifers, rivers, drains, wetlands, estuaries and coasts connect people and life that may never enter the local relationship. Spiralweb therefore treats the catchment and affected field as an outer standing around the three local readings, not as a fourth score that can be averaged with them.

If the wider effect could be material or hard to reverse and the evidence is insufficient, the affected-field standing remains Unknown. The next movement must then narrow, wait or obtain specialist and public-authority input. Spiralweb may decline to support an otherwise locally legitimate choice when credible evidence shows a basin-negative effect or a legal or ecological constraint. This is not a judgment on the steward. It is a boundary on the association’s own participation.

Absent people and living systems are not assumed to have consented. Their standing may be made visible through affected water users, legitimate local institutions, basin authorities, monitoring evidence, law, or recognised guardianship arrangements. Spiralweb does not appoint itself to speak for a river. It records whose standing is present, whose is represented, whose is absent, and when absence requires pause.

How the association may pause or decline support

The following route is an interim procedural proposition for testing. It governs Spiralweb’s own participation; it does not govern a steward’s independent choices, create authority over a place or claim to be a field-validated protocol. Until the Association adopts or delegates a more specific mandate, consequential decisions remain subject to its statutes and documented governance arrangements.

• Trigger. A pause or review is required when credible evidence indicates material or hard-to-reverse wider-field harm, legal authority is absent, affected standing is unresolved, evidence work cannot be resourced, or a conflict of interest could distort the decision.

• Temporary hold. The designated case holder or working director may place a provisional hold on Spiralweb commitments while evidence, mandate and affected standing are clarified. A temporary hold is protective, not a final judgment on the steward or proposal.

• Decision holder. A material refusal, condition or reversal must be made by the Board or an explicitly delegated and unconflicted decision body acting within a recorded mandate. AI, a funder, a coordinator or one interested individual cannot make the decision alone.

• Conflict and recusal. Anyone with a personal, financial or relational conflict must disclose it and withdraw from the decision to the extent required. The Governance and Decision Log records who participated, who recused and why the remaining holder had authority.

• Record and Return. The decision record names the evidence threshold, uncertainty, alternatives, dissent, affected interests, reasons and route to revision. The steward receives a useful explanation in an agreed language and form, with a correction route. Public detail remains bounded by consent, privacy and safety.

• Reconsideration. A decision may be reopened when evidence, legal conditions, design, affected standing or resource capacity changes. Reconsideration is not an appeal to a remote authority over the place; it is a route for revising Spiralweb’s own participation.

The hard test is procedural as well as ecological: Spiralweb must be able to refuse its own participation without making refusal arbitrary, irreversible or invisible.

A water balance proportionate to the decision

Not every field needs a complex hydrological model. Every consequential water decision does need a proportional account of source, timing, quality, demand, storage, losses, competing uses and rights. The level of measurement should rise with the consequence and irreversibility of the decision. A watering can for a small nursery and a new deep borehole do not require the same evidence.

Withdrawal, consumptive use and return flow are different. Withdrawal records water taken from a source and remains material for drawdown, access, energy, quality and rights. Consumptive use records the share not returned for reuse, commonly through evapotranspiration or incorporation in products. Return flows may support other users or ecosystems, or may return later, elsewhere or at lower quality. Plot-level irrigation efficiency therefore does not establish basin saving. The account must track the variables that the decision could materially change, rather than replace one measure with another (Grafton et al. 2018).

The knowledge architecture

Spiralweb begins from epistemic plurality without collapsing into “anything goes.” Embodied observation, inherited practice, scientific research, instruments, administrative records and professional judgment can all reveal part of a field. None is complete. Their standing depends on the question, method, provenance, scale and opportunity for correction.

The relevant distinction is not local versus scientific knowledge. It is between claims that remain answerable to their source and consequence, and claims that have been detached from them. Local knowledge can contain long memory and fine seasonal discrimination; it can also carry exclusion or be overtaken by new conditions. Science can reveal processes that are not directly sensed; it can also generalise beyond the population, climate or design actually studied. Co-production is rigorous when purpose, participants, context, methods and pathways to impact are explicit, plural and revisable rather than merely participatory in name (Tengö et al. 2014; Norström et al. 2020).

Five kinds of evidence standing

The common field inquiry

The inquiry is common; its content and outcome are not. A question may travel between fields, while observations, thresholds, authority and action remain situated. Unknown, pause, refusal and stop are legitimate movements, not failures outside the diagram. A Return may reopen the cycle and change the shared grammar.

1 Listen to place Begin with the steward’s account, existing practice, history, pressures, hopes and refusals. Ask what is alive, what is at risk and what must not be misunderstood.

2 Bound the question Name one locally useful question, the place and season it concerns, who has standing, and what falls outside the inquiry.

3 Build a baseline Record only what is needed to make the next decision responsibly. Preserve unknowns and separate current evidence from proposals.

4 Compare options Include maintenance, demand reduction, protection, assisted recovery, technical intervention and doing nothing. Compare burden, cost, reversibility, rights and downstream effects.

5 Make a small move Choose the smallest action that can produce useful learning without exceeding authority, capacity or evidence.

6 Observe consequences Use the three separate readings. Record weather, season, confounders and the limits of attribution.

7 Return and correct Return a useful account locally, revise the record, and continue, adapt, pause or stop. Share outward only within consent and privacy boundaries.

A signal, question or change begins in a living field. Flow Desk receives and bounds it without deciding its meaning in advance. Applied Field Learning supports situated inquiry, while the PG Ledger holds evidential standing, decisions, resource flows and consequences over time. Local Return allows the people and place concerned to correct, consent to, refuse or reframe the record. Only then may an agreed field record support a wider Field Return. What travels may correct not only the next field movement, but also Spiralweb’s papers, methods and governance. This is not a mandatory sequence. Any movement may narrow, pause, return, stop or remain Unknown, and no public record is required where consent or safety does not permit it.

One movement several functions

Flow Desk, Applied Field Learning, the operative formats, PG Ledger, Local Return and Field Return are not competing models. They hold different functions within one movement. Flow Desk holds and routes what arrives. Applied Field Learning shapes the inquiry. The operative formats provide proportionate forms of work. PG Ledger preserves memory and standing. Local Return keeps correction and authority close to the place. Field Return allows learning to travel without transferring that authority.

A common learning cadence

The cadence is not another stage model. It is a small set of questions that can tune different subjects to the same relational discipline while answers remain situated.

Three readings that do not compensate

A field can show ecological promise while exhausting the steward. A steward can remain committed while governance is unclear. Strong governance cannot make unsafe water safe. The three conditions are connected, but a positive reading in one cannot cancel a serious problem in another.

The stoplight method

Practice families and their limits

The literature supports several promising practice families, but effects are heterogeneous and often depend on climate, starting condition, design and management. The table below therefore treats practices as candidates to investigate, not packages to transfer.

What the external evidence supports

The evidence is strongest at the level of conditional propositions rather than universal recipes. Ostrom’s social-ecological systems framework shows why resource systems, users, governance arrangements and broader settings must be analysed together. Ecological restoration standards emphasise reference models, stakeholder engagement, measurable recovery and adaptive management. Knowledge co-production research emphasises context, plural ways of knowing, shared goals and iterative learning. The Spiralweb architecture does not conflict with these principles. That consistency is a minimum condition, not evidence that the architecture works.

Agroforestry evidence is generally promising but variable. Torralba et al. (2016) found positive average effects on biodiversity and several ecosystem services across European studies, while also finding heterogeneity and trade-offs. Krebs and Bach (2018) map permaculture principles to scientific literature; this supports component-level inquiry, not proof of a complete farming system. A 2024 paired-site study of nine Central European permaculture sites reported higher soil carbon stocks, lower bulk density and higher earthworm and plant diversity than nearby conventional fields (Reiff et al. 2024). The sites were practitioner-selected, had differing histories, and were observational pairs rather than randomised treatments. The study shows that whole-system practice can be investigated; it does not isolate causation or predict Morocco, Uganda or Mexico City.

Dryland recovery also shows that planting is not the only pathway. Farmer-managed natural regeneration in Niger works with existing rootstock and farmer decisions rather than importing a plantation model (Haglund et al. 2011; Tougiani et al. 2009). This is relevant to the inquiry grammar because it begins by asking what regenerative capacity is already present. It does not make FMNR suitable everywhere.

The literature also disciplines enthusiasm. The classic synthesis by Briske et al. (2008) found that rotational grazing was not consistently superior across rangeland experiments; it is a reference point in a continuing adaptive-management debate, not the last word. Tree planting is not a simple substitute for emissions reduction, ecosystem protection or context-specific restoration (Holl and Brancalion 2020). Water saving at plot level can fail to produce basin-level savings when consumption expands or return flows change. These are not reasons for inaction. They are reasons to make boundaries, counterfactuals and total resource use visible.

Current field pixels

The following summaries state the present evidence boundary as of 20 September 2026. They are not needs assessments, local agreements or funding decisions. Each must be returned for local correction before consequential use. This public working version uses roles rather than personal names where publication consent for this paper has not been confirmed.

Morocco as a water governance inquiry

Morocco’s recent winter rain brought important relief after a prolonged drought, but rainfall recovery does not remove structural water risk. National and basin evidence points to increasing scarcity, groundwater pressure, agricultural demand and the need to connect infrastructure with demand management and governance. FAO’s work on the Berrechid aquifer is especially relevant because it treats metering, user organisation and aquifer contracts as governance questions rather than only engineering questions. The World Bank’s Morocco climate and development work likewise frames water as a central adaptation and development constraint.

For the Berrechid aquifer, FAO reports that the annual water-balance deficit rose from about 20 million cubic metres in 2010 to about 30 million in 2016, and describes aquifer contracts, user associations and metering as parts of the response. A 2023 study of 22 groundwater samples found strong spatial variation: about 40.9 per cent were classified as excellent, 45.5 per cent as poor, 4.5 per cent as very poor and 9.1 per cent as unsuitable for human consumption (Moukhliss et al. 2023). That heterogeneity strengthens the location gate. Province-level findings do not automatically describe Had Soualem or any particular well; they become field evidence only after exact location, aquifer connection, water source and current use are confirmed.

Two Moroccan inquiries

The two Moroccan relationships share a water discipline but not the same decision. Doukkala begins with a system-mapping problem. Firmaty begins with a consequential infrastructure question. Treating them as the same kind of case would either overburden the first or under-investigate the second.

In Doukkala, the first task is to map the water field across the approximately 80-hectare holding: several wells, one large lined reservoir reported as full after the wet winter, a second smaller reservoir, existing uses and the two proposed drip-irrigated test fields. The map must show how water is sourced, stored, pumped, conveyed, lost and allocated before the test fields are treated as bounded additions. A wet winter creates present capacity; it does not by itself establish recharge, sustainable yield or future security.

The two Doukkala test fields ask one ecological question through different starting conditions. In the first, a proposed multi-strata agroforestry or food-forest design would be established within a reported state-supported pomegranate planting whose trees stand roughly six to eight metres apart, with dry, sparsely covered ground between them and limited vigour reported locally. In the second, a layered dryland agroforestry field would be designed from the beginning, including protection and fencing. Both are proposals awaiting baseline correction. They can become two documented trajectories with useful reference conditions. They are not a replicated experiment and cannot isolate the effect of one practice. Comparisons should match soil texture, slope, prior management and water supply as far as feasible and state what remains different.

The planning horizon must match living establishment. In the first year, survival is the decisive ecological indicator alongside water use, labour and reasons for loss. A six-month cycle can document preparation, planting, irrigation and early survival; it cannot demonstrate a functioning food forest or long-term water regeneration. Three to five years is a working horizon for establishment, not mature ecosystem function or regenerated hydrology. The hypothesis would be weakened if cover and biomass rise only through increasing irrigation, if soil moisture or infiltration does not improve relative to a useful reference, if water demand or labour becomes unviable, or if effects outside the test field are negative.

In Firmaty, the steward’s request to deepen an existing well brings the architecture into a real test. The field may need more reliable water to remain alive, while a deeper well may increase abstraction, reach another water-bearing layer, change quality or cost, or shift effects beyond the field. The inquiry begins with commitment to the steward and the existing permaculture practice. It does not presume that deepening is wrong, or that an alternative technology will be sufficient. Spiralweb is not deciding whether people in dry regions may use groundwater. It is deciding what it can responsibly help make possible in this particular relationship.

The Green Hub inquiry should therefore compare a locally realistic water portfolio: well rehabilitation or deepening; pumping and irrigation changes; storage and rainwater capture; safe reuse; runoff spreading and infiltration where lawful and hydrogeologically appropriate; soil-water retention, shade and crop or area adjustment; and atmospheric water harvesting. Fog collectors, radiative dew surfaces, active condensation and sorbent systems remain candidates only. Without a local yield, area, energy, water-quality, maintenance and cost calculation, they should be treated as possible household, nursery or propagation supplies rather than whole-field irrigation. Morocco’s fog-harvesting work in the Anti-Atlas shows that atmospheric capture can work in a suitable place; it does not establish suitability at Had Soualem (Fondation Dar Si Hmad, current project description). If a credible combination can reduce or remove the need to deepen the well, Spiralweb may help test it if resources become available. If not, a deeper well remains within the option set and may also be considered under lawful, technically credible and locally agreed conditions. Unknown keeps the comparison open; it is not a concealed refusal.

Firmaty also tests scale. A two-hectare green patch may hold soil, shade, habitat and practical knowledge while remaining dependent on a wider and declining water field. The inquiry should ask both how the existing practice can survive and whether its vegetation, soil cover, biomass cycles and water design reduce demand or improve the site’s capacity to hold water. More canopy, a food-forest layer or syntropic pruning is not presumed to be the answer. Each is a locally adapted design hypothesis to compare with water source, labour, maintenance, fire, species fit and the surrounding hydrology. If the field continues to dry despite responsible support, documenting that trajectory is also a valid Return.

The Spiralweb fields should not be used to make basin claims they cannot support. They can become small, well-documented pixels: how a reservoir is used after a wet winter; what a two-hectare field requires to remain viable; what changes when soil cover, irrigation, crop choice, storage or abstraction changes; and how the costs and benefits are distributed. The pixel is useful when it is both humble and precise.

Worked example: a bounded Doukkala inquiry

Xochimilco as a full field

Xochimilco demonstrates why local practice must not be treated as an empty recipient. The landscape joins food production, canals, habitat, cultural heritage, water management, markets, urban development and public authority. The chinampa-refuge model is particularly instructive because it links producer practice with canal filters and habitat refuges while keeping the social-ecological transformation visible (Eakin et al. 2019; Figueroa et al. 2022). Spiralweb can learn from this composition. It should not claim it as its own or reduce it to a demonstration site.

Water beyond administrative boundaries

Water moves across properties, municipalities, states, countries and epistemic categories. Healthy Roots contributes a conceptual language of capacity, contact, water and stewardship economy to this paper. It does not evidence an active Spiralweb field relationship. Its role here is lineage: a reminder that water may connect micro, meso and macro fields while authority remains situated.

Some legal systems have created explicit standing for water bodies and their catchments. Spain’s 2022 Mar Menor law recognises the lagoon and its basin as a legal subject represented through public, citizen and scientific bodies. Colombia’s Constitutional Court recognised the Atrato River as a subject of rights with guardianship arrangements. New Zealand’s Te Awa Tupua Act recognises the Whanganui River as a legal person. These are distinct legal and cultural arrangements, not a template and not authority for Spiralweb to speak for water. They show that governance can make the living water field and absent interests institutionally present.

From pixels to bioregional learning

A network becomes useful when difference is preserved well enough for patterns to be examined. Standardisation by fixed outcomes would erase the very conditions that matter. Total uniqueness would prevent shared learning. Spiralweb’s middle path is a common record structure with local questions, categories and thresholds.

Across fields, the network may ask questions such as:

• Which forms of cover or vegetation persist through dry periods, and at what water cost?

• When does improved irrigation reduce total demand, and when does it enable expansion?

• Which maintenance burdens fall on the steward, women, households or unpaid community labour?

• How do locally meaningful signs relate to instrument readings and institutional records?

• What conditions make a small action reversible, governable and worth continuing?

• Which questions recur across climate zones even when the answers differ?

Aggregation should therefore report patterns and exceptions, not manufacture a global score. A regional synthesis might say that water quality is repeatedly unknown, that steward labour is undercounted, or that storage is a shared constraint. It should not imply that one colour, indicator or intervention has the same meaning everywhere.

The PG Ledger and Flow Desk

The PG Ledger is the memory of the inquiry. Flow Desk is the relational operating surface that receives signals, distinguishes standing, routes attention, protects boundaries, remembers commitments and returns decisions. Together they can connect field consequence with money and governance without turning the field into a donor reporting object.

When it is a Ledger

A record becomes an active PG Ledger only when all of the following are present:

• a bounded purpose and named holder;

• legitimate authority, consent and privacy boundaries;

• dated observations and decisions with evidence standing;

• the three separate readings where relevant;

• a visible connection between resource flows and purpose;

• a locally useful Return;

• an agreed rhythm or event trigger;

• and evidence that the record can correct a later decision.

A collection of forms is not yet a Ledger. A public page is not yet a Ledger. A financial transaction list is not yet a Ledger. The defining feature is correctable continuity: later action can be traced to earlier observation, authority, decision, resource flow and consequence.

The five operative formats

The formats are not five stages, five duties or a compulsory chain. A field uses the smallest format that the decision and consequence justify. The rule is proportionality: no more legibility than consequence justifies, and no less accountability than consequence requires.

Proportional evidence and its budget

Evidence work is work. Measurement, translation, local participation, documentation and Return must be named and resourced rather than transferred silently to the steward. The Steward’s Journey governs invitation, support, pause, roles and dignity; the PG Ledger governs evidence, provenance, decision, consequence and correction. Financial design must connect the two.

Tier 3 entry gates

A proposal enters Tier 3 when any one of the following conditions is present. The tier may later be reduced only through documented evidence that removes the triggering condition; urgency, attention or available funding does not lower it.

• a new well, deeper borehole, increased pumping capacity, material increase in withdrawal, or material change in consumptive use, evapotranspiration or recoverable return flows;

• a plausible material effect on shared or hydraulically connected water, downstream users, wetlands, estuaries, future users or other life;

• a contamination, salinity or public-health pathway whose consequence could be serious;

• major or hard-to-reverse earthworks, drainage change, ecosystem conversion or infrastructure;

• unresolved permits, tenure, water rights, consent, contested standing or a material conflict of interest;

• a material financial commitment or public outcome claim that depends on an ecological or hydrological effect not yet established.

Area alone does not determine the tier. Mapping reservoirs and wells across an 80-hectare holding can begin as a bounded observation task. Deepening a well on a two-hectare field may require Tier 3 evidence because consequence, connectivity and reversibility—not plot size—set the burden. If Spiralweb cannot resource the evidence its own decision requires, the answer is to narrow or pause the decision, not to make the steward carry unpaid institutional work.

Return as a delivery obligation

A public synthesis is not a Return. Before a field summary is treated as current, the Return must have an agreed recipient, form, language, author or translator, privacy boundary, budget, due date or event trigger, and correction route. One locally useful page may be enough; 6,000 words of English synthesis usually are not.

Money as part of the field

Money changes tempo, authority, expectations and labour. It is therefore not an external administrative layer. The Financial Ledger should connect formal accounts to a locally useful account of purpose and consequence while protecting sensitive personal and banking data. Expenditure in Land and Ecology does not prove ecological improvement. Payment to a steward does not prove viability. Coordination expenditure does not prove legitimate governance. Each claim needs its own evidence.

Applying the Ledger to the association itself can create a public window into how support entered, what purpose governed it, where it moved, what field decision followed, and what remained unresolved. Over time this can test whether Spiralweb practises its own commitments. It can also reveal whether the surrounding world finds the work relevant and whether the association can carry relationships without converting them into promises. This is a proposition to be tested through use, not a claim of achieved transparency.

A responsible field budget therefore separates at least three kinds of support: stable stewardship capacity; concrete field functions, materials and repair; and coordination, evidence, translation, documentation and learning. Formal accounting remains necessary. The Ledger adds the question of what the flow made possible, whom it burdened, and what later evidence changed the account.

The role of AI and the workbench

AI can help search, translate, compare documents, detect contradictions, draft alternative framings, organise evidence and maintain provenance. It can make a small association more capable of holding a complex corpus. It can also amplify error, flatten context, invent continuity, reproduce bias and create a false impression of settled knowledge.

The Spiralweb AI workbench should therefore remain bounded by five rules.

• AI may support attention and memory. It may not confer local authority or consent.

• Every consequential field claim must retain a human and source route that can be checked.

• Sensitive knowledge is minimised and protected; refusal to digitise is a legitimate outcome.

• AI-generated synthesis is labelled as synthesis and kept distinct from local testimony and measured evidence.

• No funding, stoplight or activation decision is automated. Human holders remain accountable.

Used this way, the workbench is connective tissue rather than command. Its value is not that it knows the field. Its value is that it can help the people with standing see what has been said, what conflicts, what is missing and what requires a Return.

Minimum responsible baseline

A baseline should be light enough to use and strong enough to carry the next decision. The following is a minimum set, not a universal form. Items may be combined, translated or left out with a reason.

Measurement without metric capture

Indicators should be chosen because they can correct a decision, not because they are easy to count or attractive to a funder. A small field may begin with three indicators. A borehole decision may require specialist tests and permits. A wetland collaboration may need ecological, livelihood and institutional measures. The test is practical: what observation would change what we do next?

Some conditions should remain qualitative or private. Cultural meaning, conflict, bodily experience and sacred or protected knowledge are not improved merely by being quantified. The Ledger can record that a protected domain affected a decision without collecting its contents.

The baseline also distinguishes what is required to act safely from what would be useful to learn. Safety, authority, rights and stop conditions may be gates. Additional observations may improve learning without becoming permission slips for every small movement.

Research agenda

The present corpus is conceptually rich and contains increasingly precise field records. It is not yet a complete evidence system. The following work would materially strengthen this paper and the field architecture.

What can be stated now

• Spiralweb has a coherent candidate architecture for place-based inquiry, non-compensatory review, resource traceability and correction.

• The architecture does not conflict with several established principles in social-ecological systems, knowledge co-production, adaptive management and restoration standards. This is a minimum condition, not evidence that it works.

• The field materials show real relationships and increasingly bounded questions, but not a mature cross-field Ledger cadence.

• The common land and water paper can guide preparation now, provided it remains visibly correctable and does not outrun local agreement.

What cannot yet be stated

• That the interim pause, refusal and reconsideration route is field validated, formally delegated in every case or proven robust under pressure.

• That Spiralweb interventions have produced regenerative outcomes across fields.

• That the PG Ledger is field validated or low burden in sustained practice.

• That any one practice family is the preferred solution across the network.

• That current proposals, public horizons or budgets are funded or activated.

• That local pixels support basin, national or planetary causal claims without additional evidence.

• That the association’s current refusals prove integrity under future pressure from money, urgency or reputation.

What would weaken the framework

The framework should be revised or rejected where practice shows that:

• a named technique can safely be treated as regenerative by definition;

• one composite score can replace the three readings without hiding decisive harm;

• Unknown can be treated as Yellow without increasing risk;

• Returns do not correct later decisions or are not useful to field holders;

• stewards repeatedly carry uncompensated evidence and translation work;

• local Green readings routinely conceal wider-field Red effects;

• refusal becomes impossible, arbitrary or practically unreviewable when funding, urgency, loyalty or attention arrives; or

• the Ledger records activity and money but not cost, consequence and correction.

Conclusion

Spiralweb does not need a universal regenerative recipe. It needs a shared discipline strong enough to connect places without taking them over. Land and water provide a demanding test because their relations are physical, cultural, political and economic at once. A field can be alive and still fragile. A steward can be capable and still unable to carry unpaid burden. A visible green intervention can still deepen extraction. A pause can be the most regenerative act available.

The proposed architecture begins with contact and ends with correction. It asks what is already alive, what the place and its holders wish to protect, what evidence the next decision requires, and what the smallest responsible movement might teach. It keeps ecological condition, steward viability and governance separate enough to prevent compensation and connected enough to understand consequence. It records money without allowing money to define success. It uses AI to support memory without granting it authority.

If these disciplines become a living monthly and event-based practice, pixels may begin to form a wider intelligence: not a central view that claims to know every place, but a network in which grounded differences can warn, teach and correct one another. That is the paper’s main wager. Its truth will not be settled by the paper. It will be settled, slowly, by use.

At v0.4.1, part of the paper’s credibility still rests on what it refuses to claim. Such refusals are necessary and also cheap before consequential action. The harder test begins when money, urgency, loyalty or reputation presses toward a decision and the Ledger shows what Spiralweb chose, who held authority, who recused, who carried the burden, what it cost, what was refused, how the decision was returned and what had to be corrected.

Source and evidence map

Editorial and status note

This is a working field synthesis and framework, not a systematic review, a field agreement or a final protocol. Field status is dated 20 September 2026 and should be refreshed when local corrections, agreements, evidence or funding status change. Scientific references were selected to test the main propositions and limits, not to claim comprehensive coverage. Before programme design or consequential funding decisions, relevant claims require a documented search method, specialist review where needed, and field-reviewed Returns.

Spiralweb Stewardship Association · Applied Field Learning. Published through Spiralweb Papers under human responsibility and open to correction. Publication activates no field relationship, funding flow or protocol.