Choosing a Binding Adhesive: Paste, Methyl Cellulose and PVA

Three adhesives cover almost everything a bindery bench does: wheat starch paste, methyl cellulose, and PVA. The choice is not about which one is best; it is about how reversible the job needs to be and how much working time the step actually requires. Paste and methyl cellulose both come apart again in water. PVA, once it has set, does not.

What is wheat starch paste actually good for?

Paste is cooked wheat starch, thinned with water to a working consistency. The Book and Paper Group's own recipe notes describe a properly cooked batch as shiny, translucent, and coming off the spatula in sheets, which is the practical test a bench uses before thinning it further. Cooked paste keeps about three days unrefrigerated and about a week refrigerated, so a batch made for a morning's work is not wasted if the afternoon runs long.

Paste is the adhesive of choice wherever the bond may need to be reversed: a spine lining, a hinge, a hinge repair, a watercolour plate pasted to a page. It stays flexible once dry, which matters at a joint that will flex thousands of times, and a clean water poultice will release it years later without force. That reversibility is also its limit: paste alone gives no immediate tack and takes real time to dry, which is why it is the wrong choice anywhere a bond has to be ready to handle in minutes.

What does methyl cellulose add that paste does not?

Methyl cellulose, usually mixed at 2 to 5 percent in water by weight, is a synthetic adhesive with a neutral pH and full reversibility in water. On its own it is a genuinely weak bond, weaker than paste, which sounds like a disadvantage until you remember why a conservator sometimes wants a weak join: in a repair, a joint that lets go before the original material tears is the safer failure.

Mixed into PVA, methyl cellulose does a different job entirely. The combination slows the drying time of the PVA without giving up its strength, buying extra open time on a step where PVA alone would grab too fast to position correctly. After mixing, methyl cellulose needs roughly twenty minutes to an hour to thicken to a usable consistency, so it is not a step you can improvise five minutes before the glue is needed.

When does a job actually need PVA?

PVA sets to a strong bond and, once dry, is not meaningfully reversible with water the way paste and methyl cellulose are. That permanence is exactly what a casing-in step or a heavily sized, coated cloth needs: immediate tack, water resistance, and a bond that will not creep apart under load. It is the right adhesive wherever the job will never need to be taken apart again.

The trade-off is rigidity. A joint bonded in straight PVA is strong but stiff, and a stiff joint concentrates the flexing of opening and closing into a narrow line instead of spreading it, which is why a joint glued in PVA alone tends to crack again at the same place after a few years of use. Mixing PVA half and half with paste, a combination already used on this bench for repairing a split cloth joint, keeps the immediate tack and the water resistance while giving the join back some flexibility and some reversibility.

Matching the adhesive to the step, not the material

The honest way to choose is to ask what the step needs, not what the material is. A hinge or a lining that should come apart again in twenty years: paste. A repair where the new join should fail before the old material does: methyl cellulose, alone or added to slow another adhesive. A casing-in step, or a coated cloth that will not take a straight paste, needing tack now and strength for decades: PVA, mixed with paste or methyl cellulose wherever some flexibility is worth keeping.

Mixing is not a compromise forced by having the wrong adhesive on the shelf. Half paste and half conservation-grade PVA, or PVA slowed with methyl cellulose, are working formulas precisely because no single adhesive gives both immediate strength and long-term reversibility. Choosing one pure adhesive for every job on the bench is the mistake; choosing by what the specific join has to survive is the method.

A working rule for the bench

Reach for paste first, by default, on anything that touches paper and might one day need to come apart: linings, hinges, pasted plates, repairs. Reach for methyl cellulose where the bond should stay weak on purpose, or where another adhesive needs more open time. Reach for PVA only where the join is meant to be permanent and the material will not take a water-based paste cleanly. Keep all three mixed fresh, test on an offcut before the real piece, and write down the ratio that worked, because the next job on the same material will ask the same question again.