The Intuitive Approach to Restoration:
A Tribute to Elizabeth Lawrence

by Betty Kirke

With the needed scientific approach to conservation accelerating, many new members to the field are perhaps not cognizant of, or have chosen to disregard the primarily intuitive approach of which Elizabeth Lawrence was the master. For intuition to be valid, it must be based in experience - it should be both qualitative and quantitative. Liz had extensive experience in the haute couture before she changed fields and became the costume restorer at the Costume Institute of the Metropolitan Museum of Art. This background validated her intuitive approach and served her particularly well when dealing with altered or restyled garments.

I, upon entering this same field with a somewhat similar background, could understand well what she meant when she instructed me "to look for the clues," "see the memory of the fabric," and "let the dress tell you what it wants done to it." Being a designer, it took me a while to reverse my experiences of dominating the design. One had to become subject to the object. As Liz instructed, "you must get into the mind of the designer."

One particular gown owned by the Costume Institute, a Pingat of c.1866, had been in need of restoration for some time. I chose to take it on only after completing restoration on the Dyer wedding gown made in Paris during the same period. I felt that I had gained a sufficient foundation from the research done and actual experience in reversing the various restylings of that gown. The problems of the Pingat gown, however, had nothing to do with either alteration or restyling but the partial to total destruction of the decorative overlay fabric.

Upon careful examination, except for the staining and splitting of the armpits of the bodice, I found the silk satin foundation of the dress in excellent condition . Because this was so, it was possible to place the dress over the proper underpinnings to see its shape and the fall of its seams so that a more accurate date could be determined. Since Liz's philosophy allowed for the replacement of net, preferably with a man-made fiber, the destructed silk net was to be replaced by nylon net.

My personal approach to most problems is to start with the easier parts and proceed to the more difficult, clearing away the trivial to concentrate solely on the more demanding part of the problem. Therefore, I started to replace the bottom tier of the overlay of the skirt. I would find later that this was a good choice as it would clarify one point of the designer's thinking that might not have been recognized had I started from the waist down or with the most difficult part of the restoration. This tier was mostly intact. As there were no problems in taking an accurate pattern, this began to show me some consistencies in the designer's thinking.

This portion consisted of a layer of net that had channels sewn horizontally, enclosing strips of ribbon. The net, after the ribbon was encased, was then pleated. An additional layer of pleated net was replaced over the ribboned layer. This whole tier was worked together as an individual section, then stitched at the top onto the silk skirt with the bottom floating free.
The next area undertaken was the middle section of the front skirt. Again, everything was intact. Here, however, the net was beginning to powder as I removed it from the silk. Again there were two layers, one flat shape that lay directly against the silk and another net section that was shirred, puffed and applied to the first flat net piece along the shirred stitches. The entire section was worked as an individual part and replaced against the silk when completed. The "epaulets," the shapes repeating the shoulder epaulets, were sewn through the net onto the silk. While removed from the dress they were blocked and replaced after the net section had been sewn onto the silk.

I continued to replace the net on the sides and back of the skirt. The net now was so badly destructed that I barely managed to copy the pattern before the net powdered. This, although covering much more space, was constructed in the same way as the front mid-section, sans metallic cording. Although comprised of several pieces, the whole part was treated as an individual section. There was one flat section that matched the shape of the satin skirt. Over it was the second net skirt, again shirred and puffed. The whole section was worked off the satin and reapplied. The same treatment was repeated to the "epaulets" in the back.

The next logical step was to finish the skirt. In trying to determine what the missing part or parts were that covered the top of the skirt I took note of what parts were left. There were two more elaborate "epaulets" at each side front and eight long flat pieces of piping of the self silk, each with two mitered corners, each of varying length. One group with a bow of self silk that was still attached to the center back was much shorter in length than the other. The groups had been held in some kind of order by being pinned to cotton twill tape sometime before I started. The longer group of four pieces of piping had fragments of pleated net of varying lengths, the longest of which was four inches long.

To have a four inch pleated edging did not seem logical, but I tried that anyway. It did not work. I did more research and tried every possible design treatment from the period. None worked - or so my intuition told me. So far the dress itself was not telling me anything but what not to do. After exhausting all possibilities, I decided to put the project aside for awhile.

Although I stayed away from the museum for several weeks, I continued to project slides of the gown taken on two separate occasions, long before treatment. I also made a written list of known factors hoping something would come to mind that would lead me to the solution. I did have clues:

1. the two groups of flat silk piping
2. the silk piping's mitered corners
3. the tacked position of the piping
4. the four inches or less of pleated fragments
5. a self satin bow

And I understood the mind of the designer insofar as I knew that:

1. he worked in sections off the silk satin foundation
2. he worked from the hem up
3. he used a repetition of ideas such as in the pleated and puffed parts
4. he used a gradation of sizes in these parts
5. he used variations of the same parts, such as the epaulets

I ran through this list of known factors almost every morning. Some mornings, to impress them further on my mind, I rewrote them. I also projected the slides every other day or so, hoping to see something I had overlooked. One night I saw for the first time on one of the earlier slides what appeared to be another satin bow. Intuitively, I knew this had meaning beyond there being a bow missing. One morning soon after, the solution came to mind. Intuitively I knew it was right. I checked the thought against the lists of clues and the predictables of the designer's mind. This verified the idea. The design of the back of this gown was a simulated bow.

Naturally, in order to restore the two groups of the flat piping to a position so they would appear to be a flat simulated bow meant that I could not tamper with the already tacked points of the piping to the "epaulets" or change any of the mitered corners. Therefore, these corners falling from and below the "epaulets" had to be the fixed corners for the bottom edge of the "bow" and all mitered corners had to fall within these fixed corners. These four belonged to the shorter group of piping. When the mitered corners of the group of longer pieces were placed within these, what was left between the corners and the tacked points on the "epaulets" were easily arranged into the shape of a bow.

There still remained the problem of the purpose of the pleating. I replaced the pleated fragments. They were attached on the lower long piping extending their length to twelve inches. Now when this long group was arranged, all the empty spaces left outside the bow shape were filled as well as the remainder of the open area at the top of the skirt. The shorter group of piping was sewn to net, serving as an overskirt to the above area. This was shirred and stitched to the waist in the back. The front top panel of net to which the front piping was stitched was flat. A replica of the second missing bow was made and attached .

The top of the gown was completed almost routinely. The front side panels were overlayed. The net was replaced. The job was done. All the "clues" were accounted for and the design was consistent with the thinking of the designer. The simulated bow was a repetition and a variation in size of the self-satin bow. The back was worked in section. It was worked from the bottom up. There was repetition of ideas, gradation of size and variation in size of parts.

One could say that intuition in this case is nothing more than having the faith and patience to wait for the solution of the problem to come to mind. Liz always felt that in costume restoration, unless the restorer knew how to make clothes she or he would find it difficult to be creative when that was needed in a restoration. I believe she meant that unless you have enough experiences with clothing construction and sewing, you would not have enough data in the mind to which you could match the clues or recognize the peculiarities of the designer. The real need is to be able to respond with certainty when such a coupling occurs. Thanks to Liz Lawrence I had that experience in restoring this spectacular gown. Soon after its completion it was exhibited for the first time at the Kyoto Costume Institute in Kyoto, Japan.