A weaving drawdown is more than a coloured grid. It can be reconstructed from a threading plan and a sequence of shaft lifts: find the shaft holding each warp end, then ask whether that shaft is raised for the current pick. That relationship explains why an error can repeat down a column even when the lift sequence is correct.
The short answer in our stated rising-shaft model is simple: a cell is 1 when its warp end’s shaft belongs to that pick’s raised set, and 0 otherwise. We define 1 as warp above weft on the named reference face. The numbers describe crossings, not yarn colours. A different loom or notation needs its own interpretation before the same symbols can be used.
This original paper exercise continues The Gold Standard of Yarn Crafts, alongside the existing Weaving guide. It belongs within Learning That Transfers into Life and How X Works. The small plans below are invented teaching constructions, not commercial patterns or instructions for setting up a particular loom.
eduKateSG · Yarn Crafts
Find your next learning step
Choose the question closest to your task, or read the full lesson in order.
ROUTE 1 · CHAPTERS 1–4
Define the construction
Keep crossings, shaft assignments and notation distinct.
ROUTE 2 · CHAPTERS 5–8
Build and check the grid
Reconstruct original picks and distinguish colour from structure.
ROUTE 3 · CHAPTERS 9–14
Translate plans and views
Preserve meaning across sequences, repeats and conventions.
ROUTE 4 · CHAPTERS 15–19
Diagnose and infer
Locate mismatches and test what finite observations identify.
Chapter contents
Define the construction · Chapters 1–4
1. Begin with the crossing you want to explain
2. Keep threading, lifts and drawdown distinct
Build and check the grid · Chapters 5–8
5. Define a four-pick lift plan
6. Reconstruct the first pick one end at a time
Translate plans and views · Chapters 9–14
9. Convert a stated tie-up into raised sets
10. Keep the order of picks intact
11. Inspect floats across a repeated boundary
12. Change the face without losing strand identity
Diagnose and infer · Chapters 15–19
15. Locate a changed threading assignment
16. Recognise an impossible column disagreement
17. Infer only what the observed grid supports
Review and continue · Chapters 20–22
20. Keep a readable reconstruction record
CHAPTER 1 OF 22 · Define the construction
1. Begin with the crossing you want to explain
Imagine a chart with six columns and four rows. It contains twelve 1s and twelve 0s. That count tells us something about the grid, but it does not tell us how a shaft loom would produce it. The construction question needs a relationship between warp ends, shafts and picks.
In this lesson, a warp end is one labelled strand running in the warp direction. A pick is one labelled insertion of weft. Each cell refers to the crossing of one particular end and one particular pick. Those identities remain fixed throughout the calculation.
A shaft groups ends that move together in our simplified model. If two ends are assigned to the same shaft, they have the same up-or-down state for every pick, even if their yarn colours differ. That grouping is the central constraint we will use.
The initial task is therefore causal: what plan and lift sequence explain the crossings? It differs from merely copying a visible colour motif. A dark yarn passing underneath another yarn is still dark; a 1 in our grid means a structural relationship regardless of colour.
Before calculating, label one cell in a complete sentence: this is end 3 crossing pick 2, and 1 means that warp end lies above the weft on the specified face. That sentence prevents several later ambiguities.
CHAPTER 2 OF 22 · Define the construction
2. Keep threading, lifts and drawdown distinct
The threading tells us which shaft holds each warp end. The lift plan tells us which shafts are raised for each pick. The drawdown records the resulting over-and-under relationships in the chosen representation.
These are related descriptions, but they are not interchangeable. Changing a shaft number in the threading changes which group an end follows. Changing one pick’s lift set changes the state of every end on the affected shafts for that pick. Changing only the chart’s viewing convention changes how the same structure is displayed.
A tie-up and treadling description provides another way of specifying shaft actions on an appropriate loom. For the first examples, we bypass that extra representation and supply the raised set directly. Later we will combine a fictional tie-up and treadling into a lift plan.
Schacht's own guide to weaving drafts explains these components and the lift-plan alternative. Our six-end exercise uses its own declared orientation and invented data. It does not reproduce one of the guide’s projects.
Make three headings on your paper: end-to-shaft assignments, pick-to-raised-set assignments, and resulting cells. Keeping those sections distinct lets another reader trace a surprising cell backwards to its inputs, instead of guessing that the chart simply looks wrong.
CHAPTER 3 OF 22 · Define the construction
3. Declare the orientation and symbol key
Number ends 1 through 6 from left to right in our fixed reference view. Number picks 1 through 4 from top to bottom in the printed table. That top-to-bottom order is our presentation convention, not a universal rule for reading every published draft.
Define 1 as warp above weft and 0 as weft above warp on the named reference face. Assume a raised shaft puts its assigned warp ends above the weft for that pick. Every end is assigned to exactly one of the model’s shafts.
These statements are part of the exercise. Do not infer them from dark squares alone. A publication may use filled cells for another purpose, number shafts differently or present the advancing cloth in another direction. The legend and labels determine how to read it.
For a useful diagnostic, ask what a 1 does not mean. It does not necessarily mean black yarn, the first colour in a key, a front-side knitting stitch or an instruction to move a particular lever on every loom. It means one specified crossing state here.
A learner can check the convention with a single labelled crossing before reconstructing the whole grid. If two readers disagree about that crossing’s meaning, adding more rows will multiply their disagreement. Settle the reference face, axes and key first.
Our invented threading is T = (1, 2, 3, 1, 3, 2). Thus end 1 belongs to shaft 1, end 2 to shaft 2, end 3 to shaft 3, end 4 to shaft 1, end 5 to shaft 3 and end 6 to shaft 2.
The sequence is deliberately asymmetric. A perfectly repeating 1–2–3 pattern can make a left–right reversal hard to notice in a quick demonstration. Here the repeated groups occur in different positions, so the labels have to be followed.
| Warp end | Assigned shaft | Other end sharing this shaft |
|---|---|---|
| 1 | 1 | 4 |
| 2 | 2 | 6 |
| 3 | 3 | 5 |
| 4 | 1 | 1 |
| 5 | 3 | 3 |
| 6 | 2 | 2 |
Before looking at any lift plan, predict one structural consequence. Ends 1 and 4 must have identical cell states through all picks in this simple model. So must ends 2 and 6, and ends 3 and 5.
That prediction follows from shared movement, not from visual symmetry or matching colour. It gives us an independent check later. If the completed grid contradicts it, either the reconstruction is wrong or the actual mechanism differs from the model. The check tells us where to investigate without identifying the cause automatically.
For pick 1 raise shaft 1 only. For pick 2 raise shafts 2 and 3. For pick 3 raise shafts 1 and 3. For pick 4 raise shaft 2 only. We will write these sets as {1}, {2,3}, {1,3} and {2}.
The braces mean a set of shaft identities. They do not mean a sequence of separate weft insertions. Pick 2 has two raised shafts during one pick in our model, rather than two picks whose results should be stacked together.
Each listed set also implies which model shafts are not raised. With shafts 1, 2 and 3 as the complete set, pick 3 leaves shaft 2 unraised. An omitted shaft is not an unknown in this fully specified exercise.
| Pick | Raised shafts | Unraised shafts in the model |
|---|---|---|
| 1 | 1 | 2 and 3 |
| 2 | 2 and 3 | 1 |
| 3 | 1 and 3 | 2 |
| 4 | 2 | 1 and 3 |
Now compare the two input tables. The threading varies across the width; the lift plan varies through the picks. The drawdown combines them. Keeping those dimensions separate is the key to understanding why one error can produce a column pattern while another produces a row pattern.
CHAPTER 6 OF 22 · Build and check the grid
6. Reconstruct the first pick one end at a time
For pick 1, the raised set is {1}. Examine each end’s shaft assignment in order. End 1 is on shaft 1, so its cell is 1. End 2 is on shaft 2, so its cell is 0. End 3 is on shaft 3, so its cell is 0.
End 4 is on shaft 1, giving 1. End 5 is on shaft 3, giving 0. End 6 is on shaft 2, giving 0. The complete first row is therefore 1, 0, 0, 1, 0, 0.
This is a membership test repeated across ends: does T(i) belong to the current raised set? It does not require alternating symbols automatically. Plain weave has a familiar alternation, but a supplied shaft arrangement can generate other sequences.
Ask the learner to justify end 4 without referring to the completed row. The answer should name shaft 1 and the raised set {1}. Saying it repeats end 1 can be a useful check, but the underlying relationship explains why that repetition occurs.
Then choose a deliberately awkward cell such as end 5. Its position near the right edge does not determine its state; its assignment to shaft 3 does. This prevents a learner from replacing the actual mapping with a visual rhythm that only happens to work for a simpler example.
CHAPTER 7 OF 22 · Build and check the grid
7. Complete the drawdown and audit its columns
Apply the same membership rule to the remaining picks. The completed rows are 100100, 011011, 101110 and 010001 when written in fixed end order. The spacing in a table makes each crossing easier to inspect.
| Pick | End 1 | End 2 | End 3 | End 4 | End 5 | End 6 |
|---|---|---|---|---|---|---|
| 1 | 1 | 0 | 0 | 1 | 0 | 0 |
| 2 | 0 | 1 | 1 | 0 | 1 | 1 |
| 3 | 1 | 0 | 1 | 1 | 1 | 0 |
| 4 | 0 | 1 | 0 | 0 | 0 | 1 |
Check columns independently of the row construction. Ends 1 and 4 both read 1010 down the four picks. Ends 2 and 6 both read 0101. Ends 3 and 5 both read 0110. Those equalities agree with the threading groups established before calculation.
Now check one row and one column by tracing their cells directly to both inputs. That gives two different paths through the same data. A total count of 1s would not catch every misplaced cell because one incorrect 1 could compensate for one incorrect 0 elsewhere.
The result is a logical interlacement map within the model. It does not establish the final fabric’s appearance, dimensions, durability or suitability. Those require material and making information that the binary table deliberately leaves out.
CHAPTER 8 OF 22 · Build and check the grid
8. Distinguish crossing states from colour effects
Suppose ends 1 and 4 have different colours, although they share shaft 1. Their structural columns remain identical in our model. Their visible colours can differ because the yarn identities differ. The drawdown and colour plan answer related but separate questions.
Similarly, a cell marked 0 is not an empty crossing. It means the weft lies above the warp on the named face. The weft has its own material and colour. Leaving the cell white on paper is a notation choice, not evidence that no yarn occupies the location.
Imagine changing the weft colour at pick 3 while retaining the threading and raised set. The binary row remains 101110. A colour rendering can change substantially, but the shaft-action relationship has not changed.
This distinction helps when evaluating a photograph. A visible colour block may be influenced by yarn diameter, packing, floats and surface geometry. It should not be treated automatically as a perfect record of every crossing in the underlying structure.
A useful classroom comparison uses two copies of the same binary table with different supplied colour labels. Ask which facts are unchanged and which belong to the new colour plan. No commercial pattern or culturally specific motif is needed to understand the distinction. The original mapping already provides a clear test of structural reasoning.
CHAPTER 9 OF 22 · Translate plans and views
9. Convert a stated tie-up into raised sets
For a new paper example, define two fictional treadle actions. Action A raises shafts {1,2}. Action B raises shafts {2,3}. Suppose a supplied logical model says that using both compatible actions raises every shaft included in either set.
The combined raised set is their union: {1,2} ∪ {2,3} = {1,2,3}. Shaft 2 appears in both descriptions but is still raised once. It does not cancel, and it does not become a different height because it was mentioned twice.
That is a Boolean membership operation, not ordinary arithmetic addition and not exclusive-or. Under exclusive-or, shaft 2 would disappear because it belongs to both sets, producing the wrong logical action for our stipulated union model.
This is analysis of supplied set relationships. It is not a recommendation to press multiple treadles on an actual loom. Real mechanisms and tie-ups can have constraints that the paper model does not describe. Follow the appropriate equipment instruction and competent teaching for physical operation.
Once a raised set is obtained, reconstruct the drawdown with the same threading rule as before. An all-raised set gives a row of six 1s in our six-end example. Whether that row is useful in a real textile is a different construction and design question.
A lift plan is an ordered sequence of sets. Rearranging the picks preserves how often each set occurs but changes where its crossings occur along the cloth. A frequency table cannot replace that order.
In the main example, exchange picks 2 and 3. The new rows are 100100, 101110, 011011 and 010001. Each original row still occurs once, so the total number of 1s remains unchanged. The column histories and neighbouring relationships change.
This demonstrates why a summary such as shaft 3 is raised twice is incomplete. It does not tell us whether those raises occur together or apart, nor which other shafts are raised at the same time. Both details can affect the interlacement sequence.
For a record check, preserve pick identity through every conversion. If a draft includes a repeated section, write which picks repeat and whether additional picks are included between them. An abbreviated plan needs a clear expansion rule before it can be compared with a complete drawdown.
Ask a learner to produce a changed row order with exactly the same total cell counts. Then ask what evidence would distinguish the two records. The answer should name the ordered crossings rather than another aggregate. Sequence is information that can disappear even when every total is arithmetically correct.
CHAPTER 11 OF 22 · Translate plans and views
11. Inspect floats across a repeated boundary
Within a drawdown, consecutive equal crossing states can help describe a float in the relevant direction. Define precisely whether you are following one warp end through successive picks or one weft pick across successive ends. Those are different directions of travel.
Take the abstract repeating row 1,0,0,1,1. Looking only inside one printed block, the longest visible run of 1s is two. Repeating the block joins its final 1,1 to the next block’s initial 1, giving a run of three across the boundary.
A repeat boundary is therefore part of the structure, not an automatic break in a yarn path. The same principle applies when examining a column through a repeating pick sequence. Inspect the last and first entries together when the plan says the pattern repeats continuously.
This binary run count is not a physical float length in millimetres. That would require spacing, yarn geometry and appropriate interpretation of the textile. Nor does one run count establish that a fabric will snag, wear or perform in a particular way.
Use the example as a reading check: specify the repeat, the direction, the state being counted and whether the boundary is continuous. The additional labels are short, but they prevent a correct local count from becoming an incorrect claim about the complete repeated plan.
CHAPTER 12 OF 22 · Translate plans and views
12. Change the face without losing strand identity
On the opposite face of the same simple crossing, the over-and-under relationship reverses. If warp is above weft on the named first face, weft is above warp when the relationship is described from the other face. Keeping the same abstract end and pick identities, the binary complement is 1 − D.
For the first row 100100, that complementary crossing description is 011011. This operation changes the stated face relationship. It does not automatically describe the left-to-right appearance of a physically flipped sample on a table.
A physical flip can also reverse the apparent order of ends or picks in the viewer’s image, depending on the axis of the flip. Those position mappings must be stated separately. Complementing cell values and reversing column order are different operations even when a symmetric example conceals the difference.
Use our asymmetric threading to keep them visible. Reverse the first row’s end order and obtain 001001. Complement the original and obtain 011011. Those results answer different questions.
This lesson is therefore not a rule to invert every chart when looking at the back. It is a method: identify the strand crossing, define the viewing relationship, and then account for any spatial rearrangement. Preserving identities lets a learner explain the change instead of memorising a transformation that fails on a different view.
CHAPTER 13 OF 22 · Translate plans and views
13. Translate rising and sinking descriptions cautiously
Our initial model names the shafts raised above the weft. A sinking description may instead identify shafts lowered below it. To preserve the same target crossing map in a complete idealised two-state arrangement, the lowered set is the complement of the target raised set within all model shafts.
For target raised set {1} with shafts {1,2,3}, the corresponding lowered set is {2,3}. Copying the same marked set {1} as the lowered set would produce the opposite shaft-state relationship in this abstract model.
Louët's own David loom guidance illustrates why the mechanism and notation must be read together. Its equipment-specific explanation is not a universal instruction for all looms. Our complementary-set calculation describes the stated paper model only.
The correct translation begins with the desired physical relationship, not the appearance of a filled square. Ask whether the source marks raised shafts, lowered shafts, connections or something else. Then ask what the receiving system’s symbols mean.
A useful transfer task supplies two legends with opposite conventions and one target drawdown. Have the learner write the two action descriptions and explain why they differ. If the drawdown changes unintentionally, a notation was copied before its meaning was preserved. The remedy is to translate the relationship rather than the ink.
CHAPTER 14 OF 22 · Translate plans and views
14. Rename shafts without changing the drawdown
Shaft numbers are labels. Consistently renaming them in both the threading and lift plan should preserve every cell. This provides a strong representation check.
Rename old shaft 1 as new shaft 3, old shaft 2 as new shaft 1, and old shaft 3 as new shaft 2. The threading becomes (3,1,2,3,2,1). The four raised sets become {3}, {1,2}, {3,2} and {1}.
Reconstruct pick 1. Its raised set is now {3}; ends 1 and 4 are on shaft 3, so the row remains 100100. Applying the same check to the remaining picks recovers the whole original drawdown.
Renaming only the threading while leaving the lift sets unchanged is a different action. It changes which ends respond to a command, so the resulting crossings can change. Calling both operations renumbering would hide a meaningful difference.
This is a useful way to test software or a hand transcription. A consistent change of arbitrary labels should not change the predicted structure. When it does, inspect where the translation was applied incompletely.
The conclusion concerns the logical map. Physical shaft arrangements and equipment-specific numbering still matter when using an actual loom. Mathematical label equivalence does not authorise rearranging hardware without the relevant instructions and expertise.
Return to the main example, but change end 5 from shaft 3 to shaft 2 while leaving the lift plan unchanged. Only that end’s predicted column changes, because every other end retains the same assignment.
The original end-5 column is 0110. Shaft 2 follows the column 0101, so the changed end-5 column becomes 0101. Picks 1 and 2 still agree with the original plan; picks 3 and 4 differ.
This detail matters for diagnosis. A threading mismatch need not create a visible discrepancy on every pick. When the two relevant shafts share the same state, the error is hidden in that row. A limited sample can therefore miss a real mapping difference.
Conversely, changing one pick’s raised set can alter several columns on that row. An observed row pattern and a column pattern can suggest different questions, but neither automatically proves a cause. Incorrect transcription, notation, face or observation can imitate a construction error.
For teaching, compare the expected and changed grids cell by cell. Mark which input changed and predict the affected cells before calculating them. The prediction is the evidence of understanding. Simply finding the differences after both tables have been supplied tests visual comparison rather than the underlying shaft-to-crossing relationship.
CHAPTER 16 OF 22 · Diagnose and infer
16. Recognise an impossible column disagreement
Under the main threading, ends 1 and 4 share shaft 1. Suppose a reported drawdown gives end 1 a value of 1 and end 4 a value of 0 on the same pick. Can any raised set reproduce both values within our simple model?
No. Shaft 1 is either in that pick’s raised set or outside it. Both ends must therefore receive the same binary state. The contradictory cells cannot be repaired merely by choosing a different set of raised shafts.
That is an inconsistency result within the model, not a diagnosis of a real loom. Possible explanations include an incorrect threading record, a cell transcription error, misunderstood viewing conventions or a mechanism outside the assumed one-shaft-per-end arrangement.
A useful response names the conflict precisely: the two recorded cells disagree despite a supplied shared-shaft assignment. It then identifies which evidence would resolve the conflict. Calling the whole plan bad is less useful because it does not locate the incompatible relationship.
This is an important inverse-problem boundary. Some outputs permit several input plans; others permit none under the stated assumptions. Before trying to infer an action sequence from a grid, check whether the grid is compatible with the known grouping of ends. A successful-looking answer that ignores such a contradiction has solved a different problem.
CHAPTER 17 OF 22 · Diagnose and infer
17. Infer only what the observed grid supports
If threading is known and at least one end is observed on every shaft, one complete pick can reveal the raised set in this ideal model. Read one representative end from each shaft group and include the shaft when its cell is 1.
For the main pick 3, end 1 supplies state 1 for shaft 1, end 2 supplies state 0 for shaft 2, and end 3 supplies state 1 for shaft 3. The inferred raised set is {1,3}. Other ends provide consistency checks.
If no end in the observed sample belongs to shaft 3, its action is not identified by that sample. Raising or not raising it could produce the same observed cells. An unobserved shaft should not be assigned a state simply because a complete answer looks neater.
If threading is unknown, identical column histories can be grouped, but their equality does not prove they were physically on one shaft. Two shafts could have followed the same lift sequence throughout the observed picks.
Thus a drawdown can support an economical logical representation while leaving the actual physical arrangement ambiguous. Distinguish a possible reconstruction, a minimum representation within a stipulated model and evidence about the loom that actually produced a textile. Those are different claims, even when one attractive diagram seems to answer all three.
CHAPTER 18 OF 22 · Diagnose and infer
18. Find the minimum groups for a finite grid
In a one-shaft-per-end model, different column histories cannot be assigned to the same shaft. Our four-pick grid contains three distinct column histories: 1010, 0101 and 0110. Any such representation therefore needs at least three distinct shaft-action groups for these observed picks.
Three groups are also sufficient: assign each history to one shaft and raise it on the picks marked 1. The main threading and lift plan provide an explicit construction. We have both a lower-bound argument and a matching example.
This is a minimum for the finite binary record under the stated model. It is not proof that a photographed cloth was made on exactly three shafts. Additional shafts can share identical histories, and unobserved picks can distinguish groups that looked identical in the sample.
Now imagine extending the observation with a new pick where two previously identical columns differ. Their old grouping no longer works for the enlarged record. A minimum representation can change when the information scope changes.
This gives learners a useful evidence habit: always state which rows and columns support a structural inference. A sample that adequately demonstrates one relationship may be too small to identify the complete mechanism. The answer should grow with the evidence, rather than force new observations into an earlier convenient grouping.
CHAPTER 19 OF 22 · Diagnose and infer
19. Practise on a changed threading and lift plan
Try a new paper case before reading the answer. The threading is (2,1,2,3,1). The two raised sets are {1,3} for pick A and {2} for pick B. Write both rows in the fixed end order 1 through 5.
Pick A is 0,1,0,1,1. End 1 is on shaft 2 and remains unraised; end 2 is on shaft 1 and rises; end 3 is on shaft 2 and remains unraised; end 4 is on shaft 3 and rises; end 5 is on shaft 1 and rises.
Pick B is 1,0,1,0,0. The rows are complements because the two raised sets partition the three model shafts. That complementary relationship follows from the supplied sets; it is not a rule that every pair of successive picks must satisfy.
Now rename shafts consistently by exchanging labels 1 and 2. The threading becomes (1,2,1,3,2), and the raised sets become {2,3} and {1}. The rows remain unchanged. Explain why without recomputing every cell.
Finally report only ends 1, 2 and 3. Can those observations identify shaft 3’s action? No end in that shortened sample is assigned to it. The full plan supplies the action, but the shortened observations alone do not. This last change tests whether the learner preserves the distinction between known inputs and inferred information.
Begin with the reference face, axis order and meaning of each symbol. Record the complete set of model shafts and the one-shaft-per-end assumption. Preserve any repeat and expansion rules alongside the data.
Keep the original threading and lift plan separate from the generated drawdown. If converting from tie-up and treadling, save the intermediate raised sets. An unexpected row can then be traced to either the conversion or the final membership step.
Use a small asymmetric case as a check. Include a shared-shaft column comparison, one manually reconstructed row and a consistent shaft-renaming test. These checks expose different mistakes; a single total count does not replace them.
Record what the output leaves unresolved: yarn colours, spacing, tension, finishing, equipment compatibility and actual textile performance, for example. A binary structure can be correct while a proposed physical project remains unsuitable or untested.
For a class, ask a second learner to reconstruct a few cells from the record without seeing the answer table. If they need to guess whether 1 means raised shaft or dark yarn, improve the key. The practical value of the record is that its meaning survives a change of reader. Clear notation makes creative work easier to explain, check and adapt.
Is a drawdown the same as a colour chart?
Not necessarily. Our drawdown records which strand lies above the other at each crossing. Yarn-colour information is a separate input. A publication’s own legend determines what its cells mean.
Why must ends on one shaft have identical columns?
They share the same raised-or-unraised state on every pick in the simple model. If their reported columns differ, the data or assumptions need review. The disagreement does not by itself establish which record is wrong.
Can I reverse a rising-shaft notation for every loom?
Use the actual mechanism and notation guidance. Complementary sets preserve a target relationship in our defined two-state model, but equipment-specific operation needs appropriate instruction. A paper conversion is not a universal tie-up procedure.
Does a minimum of three column groups prove three physical shafts?
No. It establishes a minimum representation for the observed finite grid under our assumptions. Additional physical shafts can have identical observed histories, and unseen picks may distinguish them.
Does this replace the existing knitting colour-chart lesson?
No. Translating colour charts between flat rows and unturned rounds addresses a different construction and working convention. Here the central relationship is grouping warp ends by shaft and combining that grouping with a lift sequence.
CHAPTER 22 OF 22 · Review and continue
22. Choose the next making and learning route
Yarn and textile structure
Return to Yarn Crafts for choosing a technique, sampling materials and keeping a useful project record. Continue with Weaving: Balanced Warp, Intentional Weft and Respectful Textile Craft for the broader relationship among interlacement, tension, edges and purpose.
Pattern representation and quantities
The existing textile patterns, symmetry and repeats lesson develops geometric transformations and repeat descriptions. It serves a different question from reconstructing shaft actions. Budgeting usable yarn across colours, swatches and held strands addresses supply quantities once the intended construction and material choices are understood.
Primary making references
Schacht's weaving-draft explanation is a manufacturer-authored introduction to threading, tie-up, treadling and lift plans. Louët's David loom note illustrates a specific sinking-shaft convention. Read each within its own equipment and notation scope; neither is the source of our original six-end grid.
Source routes checked on 22 October 2026. No named traditional motif, museum pattern or commercial project is reproduced. Continue through Learning That Transfers into Life and How X Works by explaining one unfamiliar cell from its actual construction inputs, with the convention and limits still visible.