Resetting list styles at the top of a stylesheet with list-style: none, padding: 0 and margin: 0 is a common habit, but it throws away useful defaults. Browsers apply these values to <ul> elements on purpose, which is what makes lists readable without any markup or styling. In Chromium the defaults look like this:
ul {
list-style-type: disc;
margin-block-start: 1em;
margin-block-end: 1em;
margin-inline-start: 0px;
margin-inline-end: 0px;
padding-inline-start: 40px;
}
Rather than discard those defaults globally, it is preferable to let the browser know that a default is likely unwanted whenever a class is present on the element. A scoped reset achieves that:
ul[class] {
padding: 0;
margin: 0;
list-style-type: none;
}
The default styling now survives for plain lists and disappears only when the author opts in with a class.
Where the scoped reset breaks
Specificity is the catch. Suppose a classed list needs a different list-style-type:
ul[class] {
padding: 0;
margin: 0;
list-style-type: none;
}
.list {
list-style-type: square;
}
This declaration loses, because ul[class] carries more weight. The fix could be to raise the reset selector's weight as well:
ul.list {
list-style-type: square; /* Specificity: 0, 1, 1 */
}
/* or */
.sidebar .list {
list-style-type: square; /* Specificity: 0, 2, 0 */
}
That works, but many authors are unwilling to pay the price. Keeping element names baked into selectors conflicts with a separation of concerns, and with BEM-style conventions it is a reliable source of trouble.
Zeroing out specificity with :where()
Functional pseudo-selectors such as :is() and :where() can raise or cancel specificity. :where() in particular always contributes 0 specificity, which removes the conflict entirely:
:where(ul[class]) {
list-style: none;
}
.list {
list-style: square; /* Now this works like a charm! */
}
The same property lets a library ship styles that carry no specificity at all, leaving author CSS nothing to compete against. The demo lets you remove :where() and watch the difference.




