MITRA Forum Question Details

While I wasn't involved in composing the linked document, it may help to consider the effect of solvent on viscosity of the vehicle, and how reducing viscosity impacts the envelopment of pigment by the vehicle.​ Oil paint isn't just a simple mixture and proportion, it's a homogeneous dispersion with each pigment particle enveloped in oil. Naturally, paint that has been thinned excessively would lose the orderly distribution of pigment gained through milling. Excessively thinning oil colors results in paint with a powdery appearance due to unbound pigment. It also occurs to me that a thinned vehicle could be taken up more readily by an absorbent ground, and I imagine some is even wicked away by a semi-permeable palette.

​I'm sure our scientist Moderators will correct where I am in error, but as I understand it, fresh, pure gum spirits of turpentine that has not been exposed to the air will take up oxygen more readily than old stock from an open bottle. Because of this, fresh turpentine mixed with oil paint will increase available oxygen in the paint film and support through-drying (though not as a catalytic siccative). Old turpentine that has already oxidized will not impart this effect, but will still thin paint. Stale turpentine containing oxidation products has been shown to be more irritating to the skin, as well.

Sorry for the late reply. As to the original comment. Honestly, you are right. That document was written when I was still using the fat over lean terminology on a regular basis and that section was certainly overstated. I have since come over to the PVC school of thought. I will need to edit that section to essentially say the same thing but remove the fat over lean reference, which you rightly criticize. There are other realistic reasons to thin lower layers and add less solvent to upper layers in multiple layered systems.

However, I do sort of question the idea that heavily diluted paint has the same binding properties (different than fat or lean). The concept makes sense in theory, but it does not work exactly the same in practice. If it did, you could take a given amount of oil paint, thin it with a very large amount of solvent (let’s say OMS here to avoid the turpentine issues mentioned above) and apply it as a toning layer on a ground. In theory, this should be as well bound as neat oil paint physically distributed by the brush to a very thin layer (to the degree that is likely in practice and not a theoretical, unlikely degree).

In practice, though, it is very easy to add enough solvent to oil paint and create a film that can be easily rubbed off on your finger. We see this all of the time in modern and contemporary paintings. Sometimes they cannot even be cleaned in any manner because any physical manipulation dislodges original pigment. Yes, an absorbent ground would exacerbate this making it more likely that the binder would be drawn into the ground leaving the paint underbound. This may be a major cause of the apparent discrepancy. I have, however, seen the same effect on relatively nonabsorbent grounds but have not preformed any real scientific tests. So while I completely with your take on all theoretical grounds, there is still seems to be a difference in the bonding strength between neat oil paint and oil paint thinned with solvent.

However, I still do need to remove that phrase, as it is misleading.

If the paint is exactly at its optimal PVC what you are saying probably is technically true.  It does assume that there is not even a bit of a surplus of oil, which is probably not the case. However, I would guess that the effect is not linear and that dilution has little or no deliterious effect at relatively lower radios and rather strong effects at higher. As I said, judicious dilution is probably beneficial in multiple layered systems. This may be more of a function of maintaining tooth in the lower layers and not so much an issue of fat and lean.

There was a lot of really in-depth work done of drying oil paints in the early 20th century before synthetic materials gradually supplanted them. The positive notion of controlled dilution is somewhat reinforced by early industrial recommendations for the proportions to create a basic lead white primer and house paint. I don’t have a specific volume in front of me but they almost always listed so many pounds of lead white pigment, so much linseed oil, and so much turpentine and not simply pigment and oil. Mayer has similar suggestions for an oil ground consisting of so much Dutch boy lead paste diluted with so much turp to a workable consistency.

Relating to the subject here, I have found it preferably to dilute lead white to a heavy cream consistency when using it as an oil primer as opposed to simply using the priming knife to create a smooth priming layer out of neat paint. The results of dilution were always slightly drier feeling and more appropriate as a ground layer than my early experiments applied by physical manipulation alone. That is not to say that those early grounds were not useful, they could be very smooth and pristine, they just did not take paint as well as those thinned to a degree and applied with the priming knife.

It does seem to me that the amount of solvent that is used to create the "initial wash" is often extreme. It would be far better to dilute the paint to to a small degree and then physically spread the paint with a brush (or even better a rag as long as it is disposed of appropriately to avoid spontaneous combustion) I think that some painters believe that their later applications of more robust paint will consolidate the initial powdery layer. There may be a germ of truth in that, but for the most part, this practice is flirting with possible paint delamination. If one creates a wash and then paints on it while it is still wet, the problems you mention would be avoided.